A method and system for monitoring the health of a marine battery module

Abstract

The application relates to the technical field of battery evaluation, and particularly discloses a ship battery module health monitoring method and system, infrared image data of a battery module in a power supply process is collected; the infrared image data is subjected to time domain difference processing, an infrared flicker difference frequency characteristic value is calculated according to the pixel gray scale change degree between adjacent frames, and the infrared flicker difference frequency characteristic value is used for evaluating the periodic stability of the power supply process; by analyzing the spectral energy distribution of the infrared image data under different wave bands, a spectral distribution difference characteristic value is calculated, and the spectral distribution difference characteristic value is used for evaluating the power supply sufficiency; the infrared flicker difference frequency characteristic value and the spectral distribution difference characteristic value are input into an infrared state judgment model, the power supply state of the infrared is judged according to the model output; early warning processing is carried out, the application realizes all-weather and high-precision monitoring of the power supply state of the battery module, improves the accuracy and response speed of infrared abnormality identification, and has good safety protection effect.

Description

Technical Field

[0001] This invention relates to the field of battery evaluation technology, specifically to a method and system for monitoring the health of ship battery modules. Background Technology

[0002] With the rapid development of marine equipment, the safety and intelligence level of battery modules are receiving increasing attention. A battery module is an intermediate energy storage unit composed of multiple individual battery cells connected in series and parallel, typically serving as a fundamental component of a battery system. However, existing methods still suffer from low identification accuracy and response lag when facing complex environmental interference and various types of anomalies, making it difficult to meet the application requirements of high-reliability scenarios.

[0003] The existing technology has the following shortcomings:

[0004] Most existing infrared monitoring technologies rely on a single feature to determine infrared status, lacking the ability to model the dynamic changes in infrared radiation in multiple dimensions, making it difficult to accurately distinguish between normal fluctuations and true anomalies. Most solutions do not conduct in-depth analysis of the periodic stability and thermal radiation distribution characteristics of infrared power supply, leading to frequent misjudgments or missed detections under complex operating conditions. Furthermore, existing systems generally lack effective early warning and grading mechanisms, failing to implement differentiated alarm and intervention measures based on the severity of anomalies, thus limiting their practical application in health monitoring. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for monitoring the health of ship battery modules, so as to solve the problems mentioned above.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A method for monitoring the health of a ship battery module includes:

[0008] Collect infrared image data of the battery module during the power supply process;

[0009] Infrared image data is subjected to temporal differential processing. Based on the degree of pixel grayscale change between adjacent frames, infrared scintillation differential frequency characteristic values ​​are calculated to evaluate the periodic stability of the power supply process.

[0010] By analyzing the spectral energy distribution of infrared image data in different bands, the characteristic values ​​of spectral distribution differences are calculated to assess the adequacy of power supply.

[0011] The infrared scintillation differential frequency characteristic value and the spectral distribution difference characteristic value are input into the infrared state judgment model. Based on the model output, the power supply status of the infrared system is judged, including normal power supply status and abnormal power supply status.

[0012] Early warning processing is carried out based on abnormal power supply status.

[0013] As a further aspect of the present invention: the evaluation of the periodic stability of the power supply process specifically includes:

[0014] Infrared image data of the battery module is collected in real time during the power supply process. The infrared image data is processed by time-domain differential processing. Based on the difference in pixel grayscale values ​​between adjacent frames, the infrared flicker differential frequency characteristic value is calculated. It is determined whether the infrared flicker differential frequency characteristic value is greater than or equal to a preset threshold. If it is, the infrared signal during the power supply process is unstable; otherwise, the infrared signal during the power supply process is stable.

[0015] As a further aspect of the present invention: the process for obtaining the infrared scintillation differential frequency characteristic value is as follows:

[0016] The system collects a continuous sequence of infrared images of the battery module during the power supply process, performs grayscale processing on the images, and extracts the temporal grayscale change information of the infrared region.

[0017] Within a preset infrared region of interest, the pixel grayscale difference between adjacent frames is calculated to obtain a difference image matrix;

[0018] Each frame of the difference image is averaged in the infrared region to obtain a one-dimensional time series signal for subsequent frequency domain analysis.

[0019] The fast Fourier transform algorithm was used to perform spectral analysis on the obtained one-dimensional time series signal to extract the main frequency components of infrared scintillation.

[0020] Within the set range of normal infrared scintillation frequencies, the total energy within the corresponding frequency band is calculated, and the average energy within the corresponding frequency band is calculated using the mean calculation expression. The ratio of the average energy to the total energy is then calculated to obtain the infrared scintillation differential frequency characteristic value.

[0021] As a further aspect of the present invention: the assessment of power supply adequacy specifically includes:

[0022] Infrared image data of the battery module is collected in real time during the power supply process. By analyzing the degree of abnormality in the spectral energy distribution of the infrared image data in different bands, the characteristic value of spectral distribution difference is calculated. It is determined whether the characteristic value of spectral distribution difference is greater than or equal to a preset threshold. If it is, the infrared signal during the power supply process is insufficient; if not, the infrared signal during the power supply process is sufficient.

[0023] As a further aspect of the present invention: the process for obtaining the spectral distribution difference characteristic values ​​is as follows:

[0024] Infrared image data is acquired using multispectral imaging technology. Within a selected band range, the average spectral energy of each band is calculated to obtain the average spectral energy sequence for each band.

[0025] Discrete wavelet transform is applied to the average spectral energy sequence of each band to obtain wavelet coefficients at different scales, and the spectral distribution difference characteristic values ​​are calculated based on the wavelet coefficients.

[0026] As a further aspect of the present invention: the construction process of the infrared state judgment model is as follows:

[0027] A comprehensive feature vector is constructed by combining the infrared scintillation differential frequency feature value and the spectral distribution difference feature value. This vector serves as the input to the infrared state judgment model, minimizing the error between the predicted infrared state score and the actual infrared state score. This vector is used as the training objective. Based on the trained infrared state judgment model, the infrared state score is output. The infrared state judgment model is a random forest model.

[0028] As a further aspect of the present invention: the training process of the infrared state judgment model is as follows:

[0029] An infrared state judgment model is established using the random forest algorithm. The model consists of multiple decision trees, each of which is trained based on a randomly selected subset of samples and features to enhance the model's generalization ability. During training, the actual infrared state scores, which are manually labeled, are used as labels, and the infrared state scores predicted by the model are used as targets. The optimization objective is to minimize the mean squared error. After the model training is completed, the comprehensive feature vector extracted in real time is input into the trained random forest model to output the infrared state scores.

[0030] As a further aspect of the present invention: the step of determining the power supply status of the infrared sensor based on the model output specifically includes:

[0031] Determine whether the infrared status score during the power supply process is greater than or equal to a preset threshold. If yes, it is recorded as a normal power supply state; otherwise, it is recorded as an abnormal power supply state.

[0032] As a further aspect of the present invention: the early warning processing based on abnormal power supply status specifically includes:

[0033] The abnormal power supply time is obtained, and the abnormal power supply time is multiplied with the infrared status score to calculate the grade score. If the grade score is less than or equal to the preset first-level abnormal score, it is recorded as a first-level abnormality, triggering an on-site audible and visual alarm and simultaneously pushing a warning information to the user terminal. If the grade score is greater than or equal to the first-level abnormality score, it is recorded as a second-level abnormality. In addition to performing the first-level abnormality operation, the linkage control electric valve automatically shuts off the battery module, and at the same time sends an alarm notification to the remote management platform.

[0034] A ship battery module health monitoring system, used to implement the above-mentioned ship battery module health monitoring method, includes:

[0035] The data acquisition module collects infrared image data of the battery module during the power supply process;

[0036] The infrared stability evaluation module performs temporal differential processing on infrared image data and calculates infrared flicker differential frequency characteristic values ​​based on the degree of pixel grayscale change between adjacent frames, which are used to evaluate the periodic stability of the power supply process.

[0037] An infrared power supply adequacy assessment module analyzes the spectral energy distribution of infrared image data in different bands and calculates the spectral distribution difference characteristic value to assess power supply adequacy.

[0038] The infrared power supply status assessment module inputs the infrared scintillation differential frequency characteristic value and the spectral distribution difference characteristic value into the infrared status judgment model. Based on the model output, it judges the infrared power supply status, including normal power supply status and abnormal power supply status.

[0039] An abnormality alarm module is provided, which performs early warning processing based on abnormal power supply conditions.

[0040] The beneficial effects of this invention are:

[0041] (1) This invention constructs a composite feature extraction mechanism for infrared power supply status identification by integrating multimodal visual features in the time and spectral domains. Specifically, in terms of infrared stability assessment, based on temporal differential processing and fast Fourier transform analysis of continuous image frames, infrared flicker differential frequency feature values ​​are extracted. These feature values ​​can reflect the periodic fluctuation characteristics of infrared brightness over time and are highly sensitive to phenomena such as severe infrared jitter and irregular flickering under abnormal power supply conditions. In terms of power supply adequacy assessment, multispectral imaging technology is introduced to collect infrared radiation information in multiple bands and extract spectral distribution difference feature values ​​by combining discrete wavelet transform, thereby achieving accurate quantification of abnormal infrared thermal radiation distribution and effectively identifying non-ideal power supply conditions. By constructing the above two types of features into a comprehensive feature vector and inputting it into an infrared status judgment model trained based on the random forest algorithm, the system can maintain good discrimination ability under complex power supply environments, significantly improving the accuracy, robustness, and adaptability of infrared status identification. This method breaks through the limitations of traditional single feature analysis in dynamic infrared monitoring.

[0042] (2) This invention employs an infrared state judgment model based on the random forest algorithm, which fully explores the nonlinear mapping relationship between infrared image features and power supply status, significantly improving the system's ability to model and predict complex power supply behaviors. The model uses a comprehensive feature vector composed of infrared scintillation differential frequency feature values ​​and spectral distribution difference feature values ​​as input. By integrating multiple decision trees randomly divided in the sample and feature space, it achieves high-precision classification and scoring output of infrared power supply status. Compared with traditional linear models or single discrimination methods, random forest has stronger anti-overfitting ability and generalization performance, and can maintain stable discrimination effect in actual monitoring scenarios with noise interference, data fluctuations and environmental uncertainties. The continuous infrared state score output by the model can dynamically reflect the health level of the current power supply process and make a state judgment in combination with a preset threshold. Furthermore, the abnormal alarm module performs graded evaluation based on the score decline trend, abnormal duration and its rate of change, and automatically triggers a multi-level early warning response mechanism. For Level 1 anomalies, the system executes local audible and visual alarms and pushes alarm information to the user terminal. For Level 2 and above severe anomalies, the system automatically shuts down the battery module to prevent the accident from escalating, uploads the anomaly event log to the cloud server, and sends alarm notifications to the remote management platform, forming a complete safety closed-loop control chain. This technical solution achieves intelligent management of the entire process from infrared status recognition to proactive intervention, effectively improving safety, response timeliness, and the level of intelligent operation and maintenance. Attached Figure Description

[0043] The invention will now be further described with reference to the accompanying drawings.

[0044] Figure 1 This is a flowchart of a method for monitoring the health of a ship battery module according to the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Please see Figure 1 As shown, the present invention is a method for health monitoring of ship battery modules, comprising:

[0047] Collect infrared image data of the battery module during the power supply process;

[0048] Infrared image data is subjected to temporal differential processing. Based on the degree of pixel grayscale change between adjacent frames, infrared scintillation differential frequency characteristic values ​​are calculated to evaluate the periodic stability of the power supply process.

[0049] By analyzing the spectral energy distribution of infrared image data in different bands, the characteristic values ​​of spectral distribution differences are calculated to assess the adequacy of power supply.

[0050] The infrared scintillation differential frequency characteristic value and the spectral distribution difference characteristic value are input into the infrared state judgment model. Based on the model output, the power supply status of the infrared system is judged, including normal power supply status and abnormal power supply status.

[0051] Early warning processing is carried out based on abnormal power supply status.

[0052] The data acquisition module collects infrared image data of the battery module during the power supply process, specifically including:

[0053] During power supply, a high-definition camera and multispectral imaging technology installed in the battery module environment are used to capture infrared images of the battery module in real time. The infrared camera is deployed above the side of the battery module and equipped with an optical filter to enhance the recognition accuracy of the infrared area. The image acquisition frequency is no greater than or equal to 30 frames / second to ensure the integrity of the infrared dynamic change information. The acquired raw image data is then processed by white balance correction, color space conversion and noise reduction before being transmitted to the subsequent analysis module for extracting infrared morphological features and power supply status assessment.

[0054] In the infrared stability evaluation module, the infrared image data undergoes temporal differential processing. Based on the degree of pixel grayscale change between adjacent frames, the infrared flicker differential frequency characteristic value is calculated to evaluate the periodic stability of the power supply process. Specifically, this includes:

[0055] Infrared image data of the battery module is collected in real time during the power supply process. The infrared image data is processed by time domain difference. The infrared flicker differential frequency characteristic value is calculated based on the pixel grayscale difference between adjacent frames. It is determined whether the infrared flicker differential frequency characteristic value is greater than or equal to the preset threshold. If it is, the infrared of the power supply process is unstable. If not, the infrared of the power supply process is stable.

[0056] The process for obtaining the infrared scintillation differential frequency characteristic value is as follows:

[0057] The system collects a continuous sequence of infrared images of the battery module during the power supply process, performs grayscale processing on the images, and extracts the temporal grayscale change information of the infrared region.

[0058] Within a preset infrared region of interest, the pixel grayscale difference between adjacent frames is calculated to obtain a difference image matrix;

[0059] Each frame of the difference image is averaged in the infrared region to obtain a one-dimensional time series signal for subsequent frequency domain analysis.

[0060] The fast Fourier transform algorithm was used to perform spectral analysis on the obtained one-dimensional time series signal to extract the main frequency components of infrared scintillation.

[0061] The formula for calculating frequency domain energy is as follows:

[0062] ;

[0063] In the formula, Represents frequency The corresponding complex spectral coefficients, Indicates the number of frequencies. This is a one-dimensional time series signal obtained from the difference between adjacent frames. Represents the Fast Fourier Transform. Indicates the collection time;

[0064] The infrared pattern here reflects temperature. Infrared flickering is caused by temperature changes, which in turn are due to the periodic variation of the alternating current supplying the circuit. Within the set normal infrared flickering frequency range, the total energy within the corresponding frequency band is calculated, and the average energy within that band is calculated using the mean calculation expression. The ratio of the average energy to the total energy is then used to obtain the infrared flicker differential frequency characteristic value. The expression for calculating the total energy within the corresponding frequency band is as follows:

[0065] ;

[0066] In the formula, Indicates the maximum frequency. Indicates the minimum frequency. This represents the total energy.

[0067] In the infrared power supply adequacy assessment module, the spectral energy distribution of infrared image data in different bands is analyzed, and the characteristic values ​​of spectral distribution differences are calculated to assess power supply adequacy. Specifically, this includes:

[0068] Infrared image data of the battery module is collected in real time during the power supply process. By analyzing the degree of abnormality in the spectral energy distribution of the infrared image data in different bands, the characteristic value of spectral distribution difference is calculated. It is determined whether the characteristic value of spectral distribution difference is greater than or equal to a preset threshold. If it is, the infrared signal during the power supply process is insufficient; if not, the infrared signal during the power supply process is sufficient.

[0069] The process for obtaining the spectral distribution difference characteristic values ​​is as follows:

[0070] Infrared image data is acquired using multispectral imaging technology. Within a selected spectral range, the average spectral energy of each spectral band is calculated, resulting in an average spectral energy sequence for each band. The expression for calculating the average spectral energy is as follows:

[0071] ;

[0072] In the formula, Represents the average spectral energy. Represents the position coordinates of the image. Indicates the first The pixel grayscale value of each band. Indicates the preset area. Indicates the number of bands;

[0073] Applying discrete wavelet transform to the average spectral energy sequence of each band yields wavelet coefficients at different scales. Based on these wavelet coefficients, the spectral distribution difference characteristic values ​​are calculated using the following expression:

[0074] ;

[0075] In the formula, Indicates the characteristic value of spectral distribution difference. Indicates the number of discrete wavelet decomposition layers. Indicates the position of the discrete wavelet decomposition. This represents the wavelet coefficients.

[0076] In the infrared power supply status assessment module, the infrared scintillation differential frequency characteristic value and the spectral distribution difference characteristic value are input into the infrared status judgment model. Based on the model output, the infrared power supply status is judged, including normal power supply status and abnormal power supply status, specifically including:

[0077] A comprehensive feature vector is constructed by combining the infrared scintillation differential frequency feature value and the spectral distribution difference feature value. This vector serves as the input to the infrared state judgment model, minimizing the error between the predicted infrared state score and the actual infrared state score. This vector is used as the training objective. Based on the trained infrared state judgment model, the infrared state score is output. The infrared state judgment model is a random forest model.

[0078] It should be noted that this invention significantly improves the accuracy and robustness of infrared anomaly identification by extracting infrared scintillation differential frequency feature values ​​and spectral distribution difference feature values, performing multi-dimensional quantitative analysis of infrared power supply status from both the time and spectral domains. In infrared stability assessment, spectral analysis of the infrared grayscale time-series signal is performed based on Fast Fourier Transform (FFT) to calculate the energy distribution within a set frequency range. The ratio of average energy to total energy is introduced as the infrared scintillation differential frequency feature value, effectively capturing the abnormal trend of infrared periodic fluctuations. In power supply adequacy assessment, multispectral imaging technology is used to obtain infrared radiation information under different bands, constructing an average spectral energy sequence and combining wavelet transform to extract spectral distribution difference feature values, thereby reflecting the adequacy and thermal efficiency of the power supply. These methods not only achieve highly sensitive monitoring of dynamic changes in infrared morphology but also enhance the system's ability to identify abnormal infrared states under complex operating conditions by fusing multimodal features, providing intelligent and real-time monitoring methods.

[0079] An infrared state judgment model is established using the random forest algorithm. The model consists of multiple decision trees, each of which is trained based on a randomly selected subset of samples and features to enhance the model's generalization ability. During training, the actual infrared state scores, which are manually labeled, are used as labels, and the infrared state scores predicted by the model are used as targets. The optimization objective is to minimize the mean squared error. After the model training is completed, the comprehensive feature vector extracted in real time is input into the trained random forest model to output the infrared state scores.

[0080] The step of determining the power supply status of the infrared sensor based on the model output specifically includes:

[0081] Determine whether the infrared status score during the power supply process is greater than or equal to a preset threshold. If yes, it is recorded as a normal power supply state; otherwise, it is recorded as an abnormal power supply state.

[0082] The abnormal alarm module performs early warning processing based on abnormal power supply status, specifically including:

[0083] The abnormal power supply time is obtained, and the abnormal power supply time is multiplied with the infrared status score to calculate the grade score. If the grade score is less than or equal to the preset first-level abnormal score, it is recorded as a first-level abnormality, triggering an on-site audible and visual alarm and simultaneously pushing a warning information to the user terminal. If the grade score is greater than or equal to the first-level abnormality score, it is recorded as a second-level abnormality. In addition to performing the first-level abnormality operation, the linkage control electric valve automatically shuts off the battery module, and at the same time sends an alarm notification to the remote management platform.

[0084] A ship battery module health monitoring system, used to implement the above-mentioned ship battery module health monitoring method, includes:

[0085] The data acquisition module collects infrared image data of the battery module during the power supply process;

[0086] The infrared stability evaluation module performs temporal differential processing on infrared image data and calculates infrared flicker differential frequency characteristic values ​​based on the degree of pixel grayscale change between adjacent frames, which are used to evaluate the periodic stability of the power supply process.

[0087] An infrared power supply adequacy assessment module analyzes the spectral energy distribution of infrared image data in different bands and calculates the spectral distribution difference characteristic value to assess power supply adequacy.

[0088] The infrared power supply status assessment module inputs the infrared scintillation differential frequency characteristic value and the spectral distribution difference characteristic value into the infrared status judgment model. Based on the model output, it judges the infrared power supply status, including normal power supply status and abnormal power supply status.

[0089] An abnormality alarm module is provided, which performs early warning processing based on abnormal power supply conditions.

[0090] The working principle of this invention is as follows: High-precision, all-weather monitoring of the infrared power supply status of battery modules is achieved through multimodal feature fusion and intelligent analysis technology. The system includes a data acquisition module, an infrared stability assessment module, an infrared power supply sufficiency assessment module, an infrared power supply status assessment module, and an anomaly alarm module. The data acquisition module acquires infrared images using a high-definition camera and multispectral imaging equipment, and performs preprocessing to improve image quality. The infrared stability assessment module extracts infrared scintillation differential frequency feature values ​​based on time-domain difference and fast Fourier transform to quantify the stability of infrared periodic fluctuations. The infrared power supply sufficiency assessment module uses multispectral imaging combined with wavelet transform to calculate spectral distribution difference feature values, thereby assessing whether the fuel is providing sufficient power. These two types of features are constructed into a comprehensive feature vector, which is input into an infrared status judgment model trained based on a random forest algorithm. This model achieves accurate identification of the infrared status by minimizing the mean square error between the predicted score and the manually labeled score, outputting a continuous power supply status score, and determining normal or abnormal status based on a threshold. The anomaly alarm module calculates a graded score based on the product of the anomaly duration and the state score, implementing a multi-level early warning mechanism and supporting functions such as audible and visual alarms, user push notifications, and remote alarms. This invention achieves closed-loop management from image acquisition to intelligent decision-making and then to safety response, significantly improving the safety and intelligence level of battery modules, and has good application prospects and engineering promotion value.

[0091] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0092] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0093] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0094] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0095] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for monitoring the health of a ship battery module, characterized in that, include: Collect infrared image data of the battery module during the power supply process; Infrared image data is subjected to temporal differential processing. Based on the degree of pixel grayscale change between adjacent frames, infrared scintillation differential frequency characteristic values ​​are calculated to evaluate the periodic stability of the power supply process. By analyzing the spectral energy distribution of infrared image data in different bands, the characteristic values ​​of spectral distribution differences are calculated to assess the adequacy of power supply. The infrared scintillation differential frequency characteristic value and the spectral distribution difference characteristic value are input into the infrared state judgment model. Based on the model output, the power supply status of the infrared system is judged, including normal power supply status and abnormal power supply status. Early warning processing is carried out based on abnormal power supply status.

2. The method for monitoring the health of a ship battery module according to claim 1, characterized in that, The assessment of the periodic stability of the power supply process specifically includes: Infrared image data of the battery module is collected in real time during the power supply process. The infrared image data is processed by time-domain differential processing. Based on the difference in pixel grayscale values ​​between adjacent frames, the infrared flicker differential frequency characteristic value is calculated. It is determined whether the infrared flicker differential frequency characteristic value is greater than or equal to a preset threshold. If it is, the infrared signal during the power supply process is unstable; otherwise, the infrared signal during the power supply process is stable.

3. The method for monitoring the health of a ship battery module according to claim 2, characterized in that, The process for obtaining the infrared scintillation differential frequency characteristic value is as follows: The system collects a continuous sequence of infrared images of the battery module during the power supply process, performs grayscale processing on the images, and extracts the temporal grayscale change information of the infrared region. Within a preset infrared region of interest, the pixel grayscale difference between adjacent frames is calculated to obtain a difference image matrix; Each frame of the difference image is averaged in the infrared region to obtain a one-dimensional time series signal for subsequent frequency domain analysis. The fast Fourier transform algorithm was used to perform spectral analysis on the obtained one-dimensional time series signal to extract the main frequency components of infrared scintillation. Within the set range of normal infrared scintillation frequencies, the total energy within the corresponding frequency band is calculated, and the average energy within the corresponding frequency band is calculated using the mean calculation expression. The ratio of the average energy to the total energy is then calculated to obtain the infrared scintillation differential frequency characteristic value.

4. The method for monitoring the health of a ship battery module according to claim 1, characterized in that, The assessment of power supply adequacy specifically includes: Infrared image data of the battery module is collected in real time during the power supply process. By analyzing the degree of abnormality in the spectral energy distribution of the infrared image data in different bands, the characteristic value of spectral distribution difference is calculated. It is determined whether the characteristic value of spectral distribution difference is greater than or equal to a preset threshold. If it is, the infrared signal during the power supply process is insufficient; if not, the infrared signal during the power supply process is sufficient.

5. The method for monitoring the health of a ship battery module according to claim 4, characterized in that, The process for obtaining the spectral distribution difference characteristic values ​​is as follows: Infrared image data is acquired using multispectral imaging technology. Within a selected band range, the average spectral energy of each band is calculated to obtain the average spectral energy sequence for each band. Discrete wavelet transform is applied to the average spectral energy sequence of each band to obtain wavelet coefficients at different scales, and the spectral distribution difference characteristic values ​​are calculated based on the wavelet coefficients.

6. The method for monitoring the health of a ship battery module according to claim 1, characterized in that, The construction process of the infrared state judgment model is as follows: A comprehensive feature vector is constructed by combining the infrared scintillation differential frequency feature value and the spectral distribution difference feature value. This vector serves as the input to the infrared state judgment model, minimizing the error between the predicted infrared state score and the actual infrared state score. This vector is used as the training objective. Based on the trained infrared state judgment model, the infrared state score is output. The infrared state judgment model is a random forest model.

7. A method for monitoring the health of a ship battery module according to claim 6, characterized in that, The training process of the infrared state judgment model is as follows: An infrared state judgment model is established using the random forest algorithm. The model consists of multiple decision trees, each of which is trained based on a randomly selected subset of samples and features to enhance the model's generalization ability. During training, the actual infrared state scores, which are manually labeled, are used as labels, and the infrared state scores predicted by the model are used as targets. The optimization objective is to minimize the mean squared error. After the model training is completed, the comprehensive feature vector extracted in real time is input into the trained random forest model to output the infrared state scores.

8. The method for monitoring the health of a ship battery module according to claim 1, characterized in that, The step of determining the power supply status of the infrared sensor based on the model output specifically includes: Determine whether the infrared status score during the power supply process is greater than or equal to a preset threshold. If yes, it is recorded as a normal power supply state; otherwise, it is recorded as an abnormal power supply state.

9. A method for monitoring the health of a ship battery module according to claim 1, characterized in that, The early warning processing based on abnormal power supply status specifically includes: The abnormal power supply time is obtained, and the abnormal power supply time is multiplied with the infrared status score to calculate the grade score. If the grade score is less than or equal to the preset first-level abnormal score, it is recorded as a first-level abnormality, triggering an on-site audible and visual alarm and simultaneously pushing a warning information to the user terminal. If the grade score is greater than or equal to the first-level abnormality score, it is recorded as a second-level abnormality. In addition to performing the first-level abnormality operation, the linkage control electric valve automatically shuts off the battery module, and at the same time sends an alarm notification to the remote management platform.

10. A ship battery module health monitoring system, used to implement the ship battery module health monitoring method as described in any one of claims 1-9, characterized in that, include: The data acquisition module collects infrared image data of the battery module during the power supply process; The infrared stability evaluation module performs temporal differential processing on infrared image data and calculates infrared flicker differential frequency characteristic values ​​based on the degree of pixel grayscale change between adjacent frames, which are used to evaluate the periodic stability of the power supply process. An infrared power supply adequacy assessment module analyzes the spectral energy distribution of infrared image data in different bands and calculates the spectral distribution difference characteristic value to assess power supply adequacy. The infrared power supply status assessment module inputs the infrared scintillation differential frequency characteristic value and the spectral distribution difference characteristic value into the infrared status judgment model. Based on the model output, it judges the infrared power supply status, including normal power supply status and abnormal power supply status. An abnormal alarm module is provided, which performs early warning processing based on abnormal power supply conditions.

Images