Intelligent balanced underwater battery health monitoring and maintenance system

Abstract

The invention discloses an intelligent balanced underwater battery health monitoring and maintenance system, which comprises a multi-mode sensing module, an edge calculation processing module, a hybrid balance module, a fault prediction and protection module and a self-adaptive communication module, the system greatly improves the monitoring precision and response speed of the underwater battery by integrating a multi-modal sensing technology, edge calculation, intelligent equalization, fault prediction, dynamic communication optimization and real-time feedback, further has real-time fault early warning and risk prediction capabilities, can ensure the safe operation of the system, prolongs the service life of the battery, improves the energy utilization efficiency, and reduces the energy consumption. Communication and battery management strategies are optimized, the overall stability and operability of the system are improved, and risks caused by system faults are reduced.

Description

Technical Field

[0001] This application relates to the field of underwater battery technology, and in particular to an intelligent equalization underwater battery health monitoring and maintenance system. Background Technology

[0002] In existing technologies, underwater battery health monitoring and maintenance systems mainly suffer from the following technical defects: First, passive balancing technology suffers from low system efficiency due to its energy dissipation characteristics and has serious thermal management problems in the underwater enclosed environment; second, although active balancing technology is theoretically more efficient, its complex circuit topology makes it difficult to adapt to harsh environments such as high pressure and high humidity underwater, resulting in a high failure rate in actual operation; third, the static threshold control strategy commonly used in current systems cannot effectively cope with the dynamic characteristic changes of battery packs under complex operating conditions, resulting in insufficient balancing accuracy and affecting the overall system performance.

[0003] The energy loss during passive balancing not only reduces system energy efficiency, but the heat generated may also cause equipment overheating protection, affecting the continuity of operation; the high failure rate of the active balancing module will directly lead to the failure of battery pack consistency management and accelerate battery capacity decay; while insufficient adaptability of the control strategy may cause safety problems such as overcharging and over-discharging, which will ultimately significantly shorten the service life of the equipment, increase system maintenance costs, and seriously affect the reliability and economy of underwater operations. Summary of the Invention

[0004] The main objective of this application is to provide an intelligent equalization underwater battery health monitoring and maintenance system to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this application provides the following technical solution: An intelligent equalization underwater battery health monitoring and maintenance system includes a multimodal sensing module, an edge computing processing module, a hybrid equalization module, a fault prediction and protection module, and an adaptive communication module; The multimodal sensing module is used to collect multiple data points, preprocess and dimensionless the multiple data points, and reorganize the processed data into a first data group, a second data group, and a third data group. The edge computing processing module is used to perform data coupling calculations; The hybrid equalization module is used in conjunction with the edge computing processing module to perform coupled calculations on the first data group, the second data group, and the third data group, thereby obtaining the mode adjustment coefficient, analyzing the mode adjustment coefficient, and thus performing mode adjustment. The fault prediction and protection module is used to cooperate with the edge computing processing module to perform coupled calculations on the first data group, the second data group and the third data group, thereby obtaining fault warning coefficients for analysis, and thus determining whether there is a fault risk in the current mode; The adaptive communication module is used to cooperate with the edge computing processing module to perform data coupling calculation on the first data group, the second data group and the third data group, thereby obtaining the communication detection coefficient, and analyzing the communication detection coefficient to determine whether the communication is stable in the current mode. The integrated analysis module is used to couple all parameters under the first data group, the second data group and the third data group to generate integrated analysis coefficients, and analyze them. Based on the analysis results, it is determined whether the optimized underwater battery needs to issue a direct alarm. The feedback module is used to feed back various parameters and analysis results to the visualization terminal.

[0006] Preferably, the multimodal sensing module includes a multi-parameter sensing unit and a data fusion processing unit; The multi-parameter sensing unit is used to acquire multiple parameters of the underwater battery by deploying multi-parameter sensors and fiber optic sensing technology, including voltage fluctuation coefficient, current harmonic distortion rate, impedance phase angle, coulombic efficiency, temperature non-uniformity, thermal time constant, heat dissipation efficiency, thermal runaway risk index, capacity decay slope, internal resistance growth rate, polarization voltage increment, and aging consistency index. The data fusion processing unit removes noise from the acquired raw parameters, performs spatiotemporal alignment of the data, performs feature extraction and data compression, and reorganizes the data into a first data group, a second data group, and a third data group. The first data set includes voltage fluctuation coefficient A, current harmonic distortion rate B, impedance phase angle C, and coulomb efficiency D. The second set of data includes temperature non-uniformity E, thermal time constant F, heat dissipation efficiency G, and thermal runaway risk index H. The third data set includes the capacity decay slope I, the internal resistance growth rate J, the polarization voltage increment K, and the aging consistency index L.

[0007] Preferably, the hybrid equalization module includes an active equalization control unit and an intelligent switching management unit; The active equalization control unit is used to extract parameters from the first data group, the second data group, and the third data group, and input them into the edge computing processing module. The edge computing processing module performs feature fusion through a multi-layer neural network to calculate and obtain the mode adjustment coefficient. The intelligent switching management unit is used to analyze the calculated mode adjustment coefficients and determine whether a mode switch is needed based on the analysis results. The specific analysis method is as follows: when When this time, it means that the underwater battery does not need to switch modes. when When this occurs, it indicates that the underwater battery needs to switch modes.

[0008] Preferably, the edge computing processing module calculates and obtains the mode adjustment coefficient using the following formula;

[0009] In the formula: A is the voltage fluctuation coefficient, E is the temperature non-uniformity, J is the internal resistance growth rate, and L is the aging consistency index.

[0010] Preferably, the fault prediction and protection module includes an intelligent diagnosis and prediction unit and a graded protection execution unit; The intelligent diagnostic prediction unit is used to extract parameters from the first data group, the second data group, and the third data group, and input them into the edge computing processing module. The edge computing processing module performs feature fusion through a multi-layer neural network to calculate and obtain the fault warning coefficient. The hierarchical protection execution unit is used to perform data analysis on the calculated fault warning coefficients, and based on the analysis results, determine whether there is a fault risk in the current mode. The specific steps are as follows: when This indicates that there is currently no risk of failure for the underwater battery; when This indicates that the current underwater battery is at risk of failure.

[0011] Preferably, the edge computing processing module calculates and obtains the fault warning coefficient using the following formula;

[0012] In the formula: B is the current harmonic distortion rate, E is the temperature non-uniformity, H is the thermal runaway risk index, I is the capacity decay slope, J is the internal resistance growth rate, K is the polarization voltage increment, and L is the aging consistency index.

[0013] Preferably, the adaptive communication module includes a multi-mode communication control unit and a data optimization transmission unit; The multi-mode communication control unit is used to extract the first data group, the second data group and the third data group, including inputting the extracted parameters into the edge computing processing module, and performing feature fusion through a multi-layer neural network by the edge computing processing module to calculate and obtain the communication detection coefficient; The data optimization and transmission unit is used to analyze the calculated data and determine whether the communication is stable in the current mode based on the analysis results. The specific steps are as follows: when This indicates that the underwater battery signal transmission is currently unstable. when At this time, it indicates that the underwater battery needs to send a stable signal.

[0014] Preferably, the edge computing processing module calculates the communication detection coefficient using the following formula:

[0015] In the formula: A is the voltage fluctuation coefficient, B is the current harmonic distortion rate, C is the impedance phase angle, D is the coulomb efficiency, E is the temperature non-uniformity, F is the thermal time constant, G is the heat dissipation efficiency, H is the thermal runaway risk index, K is the polarization voltage increment, I is the capacity decay slope, and L is the aging consistency index.

[0016] Preferably, the integrated analysis module includes parameters from the first data group, the second data group, and the third data group, and inputs the extracted parameters into the edge computing processing module. The edge computing processing module performs feature fusion through a multi-layer neural network to calculate and obtain the integrated analysis coefficients.

[0017] In the formula: TJX is the mode adjustment coefficient, GZY is the fault warning coefficient, and CDC is the communication detection coefficient; The integrated analysis module analyzes the calculated data and, based on the analysis results, determines whether the optimized underwater battery needs to trigger an immediate alarm. The specific method is as follows: when When this time, it means that the underwater battery does not need to trigger a direct alarm. when When this occurs, it indicates that the underwater battery needs to issue an immediate alarm. Attached Figure Description

[0018] Figure 1 This is the system flowchart for this application.

[0019] In the diagram: 1. Multimodal sensing module; 11. Multi-parameter sensing unit; 12. Data fusion processing unit; 2. Edge computing processing module; 3. Hybrid equalization module; 31. Active equalization control unit; 32. Intelligent switching management unit; 4. Fault prediction and protection module; 41. Intelligent diagnosis and prediction unit; 42. Hierarchical protection execution unit; 5. Adaptive communication module; 51. Multimodal communication control unit; 52. Data optimization transmission unit; 6. Integrated analysis module; 7. Feedback module. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0021] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationships and movements between components in a specific orientation (as shown in the figures). If the specific orientation changes, the directional indications also change accordingly. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0022] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0023] Example 1: Please refer to Figure 1 A smart equalization underwater battery health monitoring and maintenance system includes a multimodal sensing module 1, an edge computing processing module 2, a hybrid equalization module 3, a fault prediction and protection module 4, and an adaptive communication module 5. The multimodal sensing module 1 is used to collect multiple data points, preprocess and dimensionless the multiple data points, and reorganize the processed data into a first data group, a second data group, and a third data group. Edge computing processing module 2 is used to perform data coupling computation; The hybrid equalization module 3 is used in conjunction with the edge computing processing module 2 to perform coupled calculations on the first data group, the second data group, and the third data group, thereby obtaining the mode adjustment coefficients, analyzing the mode adjustment coefficients, and thus performing mode adjustment. The fault prediction and protection module 4 is used to cooperate with the edge computing processing module 2 to perform coupled calculations on the first data group, the second data group and the third data group, thereby obtaining fault warning coefficients for analysis, and thus determining whether there is a fault risk in the current mode; The adaptive communication module 5 is used to cooperate with the edge computing processing module 2 to perform data coupling calculation on the first data group, the second data group and the third data group, thereby obtaining the communication detection coefficient, and analyzing the communication detection coefficient to determine whether the communication is stable in the current mode; The integrated analysis module 6 is used to couple all parameters under the first data group, the second data group and the third data group to generate integrated analysis coefficients, and analyze them. Based on the analysis results, it is determined whether the optimized underwater battery needs to be directly alarmed. Feedback module 7 is used to feed back various parameters and analysis results to the visualization terminal.

[0024] In this embodiment, different types of sensor data are collected through the multimodal sensing module 1. Advanced preprocessing and dimensionless conversion techniques are used to organize the data into a first data group, a second data group, and a third data group. This process not only improves the diversity of data acquisition but also removes noise and nonlinear factors, ensuring the accuracy of data processing.

[0025] Edge computing module 2 employs advanced coupled computing technology to process multiple data sets in real time, thereby achieving efficient analysis of battery health status. Edge computing significantly reduces data transmission latency and improves the speed of real-time data processing. This is particularly important for the health monitoring of underwater battery systems, especially in complex underwater environments, enabling rapid response and decision-making and avoiding the latency issues associated with central server computation.

[0026] The hybrid equalization module 3 works in conjunction with the edge computing processing module to perform coupled calculations on multiple data sets, generate mode adjustment coefficients, and adjust the mode based on the calculation results. Through intelligent equalization mode adjustment, the system can automatically adjust the battery's working state based on real-time data, optimize the battery's charging and discharging efficiency, extend the battery's lifespan, and improve the battery's energy utilization efficiency.

[0027] The fault prediction and protection module 4 works in collaboration with the edge computing module to obtain fault warning coefficients through coupled computing and conduct risk assessments of the battery system. This module can detect potential fault risks in advance, predict abnormal states that may occur in the battery system, and take timely measures to prevent them, avoid equipment damage or safety hazards, and ensure the stability and safety of the system.

[0028] The adaptive communication module 5 collaborates with the edge computing module to perform data coupling calculations using multiple data groups, obtains communication detection coefficients, and determines whether the current communication mode is stable based on the analysis results. This module can monitor and optimize underwater communication quality in real time, dynamically adjust communication parameters in complex underwater environments, ensure the reliability and stability of data transmission, avoid communication interruptions or data loss, and ensure continuous interaction between the system and the terminal.

[0029] The integrated analysis module 6 performs coupled analysis on all parameters, generates integrated analysis coefficients, and determines the battery health status. It triggers alarms when necessary. By integrating multiple data and analysis results, this module can provide a more accurate and comprehensive battery health assessment and promptly alert when potential battery faults are detected, effectively preventing accidents and guiding subsequent maintenance and adjustments.

[0030] Feedback module 7 transmits various parameters and analysis results to the visualization terminal, providing a real-time monitoring interface. Through visualization, operators can intuitively understand the real-time status of the battery and the analysis results, facilitating decision-making and taking necessary actions. Simultaneously, it improves the user experience and decision-making efficiency, effectively avoiding human error or operational delays.

[0031] By integrating multimodal sensing technology, edge computing, intelligent balancing, fault prediction, dynamic communication optimization, and real-time feedback, the system greatly improves the monitoring accuracy and response speed of underwater batteries. This enables real-time fault warning and risk prediction capabilities, ensuring safe system operation, extending battery life, improving energy utilization efficiency, optimizing communication and battery management strategies, enhancing the overall stability and operability of the system, and reducing the risks caused by system failures.

[0032] Example 2: Please refer to Figure 1 The multimodal sensing module 1 includes a multi-parameter sensing unit 11 and a data fusion processing unit 12; The multi-parameter sensing unit 11 is used to acquire multiple parameters of the underwater battery by deploying multi-parameter sensors and fiber optic sensing technology, including voltage fluctuation coefficient, current harmonic distortion rate, impedance phase angle, coulombic efficiency, temperature non-uniformity, thermal time constant, heat dissipation efficiency, thermal runaway risk index, capacity decay slope, internal resistance growth rate, polarization voltage increment, and aging consistency index. The data fusion processing unit 12 removes noise from the acquired raw parameters, performs spatiotemporal alignment of the data, performs feature extraction and data compression, and reorganizes the data into a first data group, a second data group, and a third data group. The first data set includes voltage fluctuation coefficient A, current harmonic distortion rate B, impedance phase angle C, and coulomb efficiency D. The second set of data includes temperature non-uniformity E, thermal time constant F, heat dissipation efficiency G, and thermal runaway risk index H. The third data set includes the capacity decay slope I, the internal resistance growth rate J, the polarization voltage increment K, and the aging consistency index L.

[0033] In this embodiment, the multimodal sensing module 1, by deploying a multi-parameter sensing unit 11 and fiber optic sensing technology, can acquire multiple key parameters of the underwater battery, such as voltage fluctuation coefficient, current harmonic distortion rate, and impedance phase angle. These parameters comprehensively reflect the electrical and thermal state of the battery. Compared to traditional single-sensor solutions, integrating multiple sensing technologies can significantly improve the system's monitoring accuracy and range. This diversified monitoring approach can capture the battery's operating status in a timely and comprehensive manner, ensuring a more accurate assessment of the battery's health. Especially in complex underwater environments, these parameters can provide more information to help identify potential faults.

[0034] The data fusion processing unit 12 employs advanced data processing algorithms to suppress noise and perform spatiotemporal alignment on the raw data acquired by the sensors, as well as feature extraction and data compression. The processed data is then reorganized into three data groups to facilitate subsequent analysis and decision-making. Noise filtering and spatiotemporal alignment eliminate the impact of environmental factors and sensor biases on data accuracy. Feature extraction and data compression further reduce the data volume, improve data processing efficiency, and ensure efficient data transmission and storage. This processing method reduces computational resource consumption and improves system response speed, making it particularly suitable for resource-constrained underwater environments.

[0035] Example 3: Please refer to Figure 1 The hybrid equalization module 3 includes an active equalization control unit 31 and an intelligent switching management unit 32; The active equalization control unit 31 is used to extract parameters from the first data group, the second data group and the third data group, and input them into the edge computing processing module 2. The edge computing processing module 2 performs feature fusion through a multi-layer neural network to calculate and obtain the mode adjustment coefficient. The intelligent switching management unit 32 is used to analyze the calculated mode adjustment coefficients and determine whether a mode switch is needed based on the analysis results. The specific analysis method is as follows: when When this time, it means that the underwater battery does not need to switch modes. when When this occurs, it indicates that the underwater battery needs to switch modes.

[0036] In this embodiment, the active balancing control unit 31 extracts parameters from the first, second, and third data sets and inputs these data into the edge computing processing module 2 for processing. The edge computing module employs a multi-layer neural network to perform feature fusion on the data, thereby calculating and obtaining the mode adjustment coefficient. Through feature fusion and calculation using the multi-layer neural network, the system can comprehensively consider multiple battery health parameters, improving the accuracy of the mode adjustment coefficient. This deep learning-based feature fusion method not only improves the intelligence level of decision-making but also enables adaptive adjustment, fully responding to the complex and ever-changing operating conditions of underwater batteries.

[0037] The intelligent switching management unit 32 analyzes the calculated mode adjustment coefficient to determine whether a mode switch is necessary. The specific switching logic is as follows: when the mode adjustment coefficient THX ≤ 0.65, the battery does not need to switch modes; when THX > 0.65, the battery needs to switch modes. This judgment mechanism is based on threshold control, ensuring that the system adjusts the operating mode in real time according to the current battery state, avoiding frequent and unnecessary mode switches, thereby reducing system energy consumption and complexity. Simultaneously, it can optimize the system based on the battery's health status, placing the battery in the most suitable operating mode, extending battery life and improving system stability.

[0038] Example 4: Please refer to Figure 1 The edge computing processing module 2 calculates and obtains the mode adjustment coefficient using the following formula;

[0039] In the formula: A is the voltage fluctuation coefficient, E is the temperature non-uniformity, J is the internal resistance growth rate, and L is the aging consistency index.

[0040] In this embodiment, the edge computing processing module 2 calculates the mode adjustment coefficient TJX by weighting multiple key battery health indicators such as voltage fluctuation coefficient A, temperature non-uniformity E, internal resistance growth rate J, and aging consistency index L. Among these, voltage fluctuation coefficient A and temperature non-uniformity E have a significant impact on the adjustment coefficient, while internal resistance growth rate J and aging consistency index L have relatively smaller impacts.

[0041] By setting different weights, this formula can prioritize key parameters based on different battery health states, ensuring that the battery adjusts its operating mode first under the influence of the most important factors. This design allows the battery to more precisely adjust for critical factors that could lead to system failure during optimized operation.

[0042] By simultaneously considering multiple parameters such as voltage fluctuations, temperature non-uniformity, internal resistance growth rate, and aging consistency index, the edge computing module can comprehensively evaluate various battery health indicators, fully reflecting the battery's current operating status. This multi-dimensional evaluation mechanism provides the system with a more comprehensive and accurate battery health analysis. Compared to judgment based on a single parameter, this comprehensive consideration not only reduces misjudgments caused by a single factor but also dynamically adjusts the battery's operating mode, thereby preventing the system from continuing to operate when a partial battery failure occurs, thus improving system stability and safety.

[0043] Example 5: Please refer to Figure 1 The fault prediction and protection module 4 includes an intelligent diagnosis and prediction unit 41 and a graded protection execution unit 42. The intelligent diagnostic prediction unit 41 is used to extract parameters from the first data group, the second data group and the third data group, including, and input them into the edge computing processing module 2. The edge computing processing module 2 performs feature fusion through a multi-layer neural network to calculate and obtain the fault warning coefficient. The graded protection execution unit 42 is used to perform data analysis on the calculated fault warning coefficients, and based on the analysis results, determine whether there is a fault risk in the current mode. The specific steps are as follows: when This indicates that there is currently no risk of failure for the underwater battery; when This indicates that the current underwater battery is at risk of failure.

[0044] In this embodiment: the intelligent diagnostic prediction unit 41 inputs various battery health parameters from the first, second, and third data groups to the edge computing processing module 2, and uses a multi-layer neural network to perform feature fusion, thereby calculating the fault warning coefficient. This process combines deep learning algorithms, enabling accurate prediction of whether the battery has potential faults based on multiple health indicators.

[0045] Through processing by edge computing module 2, the system calculates the fault warning coefficient TJX based on multiple battery health parameters. This coefficient dynamically reflects the battery's health status and provides a basis for subsequent fault risk assessment. Using deep learning technology with multi-layer neural networks, the intelligent diagnostic prediction unit can extract complex nonlinear relationships from multiple parameters, effectively improving the accuracy of fault prediction. Compared to traditional fault detection methods based on a single indicator, deep learning can combine more battery health data, reducing misjudgments caused by fluctuations in individual parameters, resulting in more accurate fault prediction and early identification of potential risks.

[0046] The graded protection execution unit 42 analyzes data based on the calculated fault warning coefficient TJX to determine whether there is a fault risk in the current battery mode. Specifically, when TJX ≤ 1, the battery is considered to have no fault risk; when TJX > 1, the battery is considered to have a fault risk. This graded judgment mechanism is simple and efficient, allowing for real-time assessment of fault risk during battery operation and avoiding unnecessary protection actions due to excessively high or low risk. When the fault risk is low, the system can operate normally, avoiding over-protection; while when the fault risk is high, the system can take timely protective measures to effectively reduce damage and ensure the safe operation of the battery and system.

[0047] Example 6: Please refer to Figure 1 The edge computing processing module 2 calculates and obtains the fault warning coefficient using the following formula;

[0048] In the formula: B is the current harmonic distortion rate, E is the temperature non-uniformity, H is the thermal runaway risk index, I is the capacity decay slope, J is the internal resistance growth rate, K is the polarization voltage increment, and L is the aging consistency index.

[0049] In this embodiment, the edge computing processing module 2 calculates the fault warning coefficient GZY using a formula that integrates multiple battery health indicators, such as current harmonic distortion rate B, temperature non-uniformity E, thermal runaway risk index H, capacity decay slope I, internal resistance growth rate J, polarization voltage increment K, and aging consistency index L. In particular, the formula incorporates several complex nonlinear factors, reflecting the intricate relationships between these parameters. By integrating multiple key battery health parameters and introducing a nonlinear calculation model, the formula can accurately capture the complex state of the battery. This method is more accurate than traditional linear models and can handle the complex effects of battery performance degradation and environmental changes, thereby improving the accuracy and early warning capability of fault warnings.

[0050] This formula places particular emphasis on the combined effect of thermal runaway risk (H) and electrical parameters. These factors collectively influence the battery's failure risk, especially when temperatures rise or the battery load is excessive, significantly increasing the failure risk. By incorporating the interaction between thermal effects and electrical parameters into the calculation, the system can more accurately predict failure risk. Particularly in underwater battery environments, where temperature and current fluctuations have a significant impact on battery performance, this comprehensive calculation enables real-time monitoring and prevention of failures caused by thermal runaway or electrical problems.

[0051] The formula includes indicators such as the internal resistance growth rate J, polarization voltage increment K, and aging consistency index L, which reflect the aging and degradation process of the battery. By fusing these parameters, the system can assess the battery's health status at different aging stages. Through accurate modeling of the battery aging process, the system can issue timely fault warnings at the onset of battery aging, rather than relying solely on obvious faults in the final stage. This predictive capability can effectively extend battery life, reduce sudden failures caused by battery aging, and optimize battery management and maintenance strategies.

[0052] Example 7: Please refer to Figure 1 The adaptive communication module 5 includes a multi-mode communication control unit 51 and a data optimization transmission unit 52; The multi-mode communication control unit 51 is used to extract the first data group, the second data group and the third data group, including inputting the extracted parameters into the edge computing processing module 2, and performing feature fusion through the multi-layer neural network of the edge computing processing module 2 to calculate and obtain the communication detection coefficient. The data optimization and transmission unit 52 is used to analyze the calculated data and determine whether the communication is stable in the current mode based on the analysis results. The specific steps are as follows: when This indicates that the underwater battery signal transmission is currently unstable. when At this time, it indicates that the underwater battery needs to send a stable signal.

[0053] In this embodiment, the multi-mode communication control unit 51 in the adaptive communication module 5 extracts key parameters from the first, second, and third data groups and inputs these data into the edge computing processing module 2. This data undergoes feature fusion via a multi-layer neural network in the edge computing module to calculate the communication detection coefficient. This data extraction and processing method ensures a comprehensive assessment of communication quality, enabling the extraction of features related to communication stability from multiple battery health and environmental parameters. Feature fusion using deep learning methods fully considers factors such as battery health, environmental changes, and current fluctuations, improving the accuracy and reliability of communication detection. This comprehensive data processing method allows the system to more accurately predict and detect communication instability.

[0054] The multi-mode communication control unit 51 uses a multi-layer neural network in the edge computing processing module to perform feature fusion on the extracted battery and environmental parameters, thereby calculating the communication detection coefficient GZY. The multi-layer neural network can effectively uncover potential nonlinear relationships in the data, especially the complex relationship between battery state and communication quality. Compared to traditional linear analysis methods, deep learning can capture more information, thus improving the accuracy of the communication detection coefficient. This not only enhances the system's ability to judge communication stability in complex environments but also enables it to adapt to various changes in real time, ensuring the stability of the communication link in the underwater environment.

[0055] The data optimization and transmission unit 52 performs data analysis based on the calculated communication detection coefficient GZY to determine the communication stability under the current mode. Specifically, the judgment rule is: when GZY ≤ 1, it indicates that the current underwater battery's communication signal is unstable; when GZY > 1, it indicates that the communication signal is stable. This judgment mechanism based on the communication detection coefficient allows the system to evaluate the communication status in real time and perform necessary optimizations or adjustments when signal instability is detected. In this way, the system can automatically determine whether the communication is in its optimal state and take corresponding measures to improve it, thereby ensuring that the underwater battery system can maintain a stable communication connection in complex environments and avoid data loss or system operation failure due to signal instability.

[0056] Example 8: Please refer to Figure 1 The edge computing processing module 2 calculates and obtains the communication detection coefficient using the following formula:

[0057] In the formula: A is the voltage fluctuation coefficient, B is the current harmonic distortion rate, C is the impedance phase angle, D is the coulomb efficiency, E is the temperature non-uniformity, F is the thermal time constant, G is the heat dissipation efficiency, H is the thermal runaway risk index, K is the polarization voltage increment, I is the capacity decay slope, and L is the aging consistency index.

[0058] In this embodiment, the edge computing processing module 2 calculates the communication detection coefficient (CDC) using a complex formula that integrates multiple battery health parameters. This formula combines several key parameters, including voltage fluctuation coefficient (A), current harmonic distortion rate (B), impedance phase angle (C), coulombic efficiency (D), temperature non-uniformity (E), thermal time constant (F), heat dissipation efficiency (G), thermal runaway risk index (H), polarization voltage increment (K), capacity decay slope (I), and aging consistency index (L). This dynamically reflects the impact of battery health on communication stability. By integrating multiple battery health and environmental parameters, this calculation method can accurately assess the impact of battery status on communication quality. Compared to judging a single factor, this comprehensive approach can more accurately capture the complex relationship between battery performance changes and communication stability. It can better handle the impact of factors such as battery aging, temperature fluctuations, and load changes on communication, improving the reliability of communication detection.

[0059] The exponential and hyperbolic tangent functions in the formula introduce nonlinear factors, enabling a more accurate simulation of the nonlinear relationship between battery health and communication quality, particularly the effects of battery temperature, capacity degradation, and polarization voltage. By introducing nonlinear calculations, the formula captures the complex and dynamic relationship between battery health and communication stability, avoiding the simplification and distortion inherent in traditional linear models. Especially under conditions of battery aging, capacity degradation, or extreme operating environments, the system can more accurately assess communication risks, predict and prevent communication quality degradation in advance, and ensure stable system operation.

[0060] The formula pays particular attention to the battery's thermal characteristics, such as temperature non-uniformity (E), thermal time constant (F), heat dissipation efficiency (G), and thermal runaway risk index (H). These factors, through exponential and constant terms, jointly affect communication quality. Thermal effects typically lead to a decline in battery performance; therefore, considering these parameters helps in accurately assessing communication quality. By taking the battery's thermal characteristics into account, the system can more accurately predict communication problems caused by battery overheating or thermal runaway. Especially in underwater or high-load environments, the impact of battery temperature on communication quality is significant. Dynamic adjustments to these environmental factors improve communication stability and ensure the reliability of the battery management system under extreme conditions.

[0061] The formula considers both the capacity decay slope (I) and the aging consistency index (L), both parameters reflecting the battery's degradation over time. Battery aging and capacity degradation directly affect its operating efficiency, thus impacting communication stability. By incorporating these aging and capacity decay parameters, the system can dynamically predict changes in communication signal quality based on actual battery usage. This effectively mitigates communication instability caused by battery aging and allows for early responses in the early stages of battery aging, extending the effective lifespan of both the battery and the system.

[0062] This real-time calculation and automatic adjustment capability enhances the system's adaptability and responsiveness, especially in dynamic underwater environments where the health of the battery and communication systems constantly changes. Through precise real-time adjustments, the system can reduce communication interruptions or data loss caused by signal fluctuations, ensuring the long-term stable operation of the battery management system.

[0063] Example 9: Please refer to Figure 1 The integrated analysis module 6 inputs the parameters from the first data group, the second data group, and the third data group into the edge computing processing module 2, and performs feature fusion through a multi-layer neural network in the edge computing processing module 2 to calculate and obtain the integrated analysis coefficients.

[0064] In the formula: TJX is the mode adjustment coefficient, GZY is the fault warning coefficient, and CDC is the communication detection coefficient; The integrated analysis module 6 analyzes the calculated data and, based on the analysis results, determines whether the optimized underwater battery needs to trigger an immediate alarm. The specific method is as follows: when When this time, it means that the underwater battery does not need to trigger a direct alarm. when When this occurs, it indicates that the underwater battery needs to issue an immediate alarm.

[0065] In this embodiment: the integrated analysis module 6 inputs the battery health parameters of the first data group, the second data group, and the third data group to the edge computing processing module 2, and performs feature fusion through a multi-layer neural network to calculate the integrated analysis coefficient THX. This coefficient is derived by a weighted combination of three key parameters, comprehensively considering the battery's health status, failure risk, and communication stability.

[0066] By comprehensively analyzing multi-dimensional battery health information, the system can fully assess battery status in complex operating environments. Combining various parameters, it considers not only battery performance but also fault prediction and communication stability, accurately reflecting the overall operating status of the battery. This diversified analysis method ensures higher fault warning accuracy, avoids false alarms or missed alarms caused by a single factor, and improves the intelligence and accuracy of battery management.

[0067] The integrated analysis coefficient THX is calculated by weighting the mode regulation coefficient TJX, fault warning coefficient GZY, and communication detection coefficient CDC. This weighting method rationally allocates weights according to the importance of each coefficient, ensuring that the final integrated analysis coefficient reflects the overall battery situation. The weighted calculation ensures that the contribution of each key parameter to the final evaluation result is commensurate with its importance. For example, if the battery is in a high-fault-risk state, the fault warning coefficient GZY may receive a higher weight, thus making the system focus more on fault warning rather than solely relying on communication quality. This flexible weighting method enhances the system's adaptability and intelligent decision-making capabilities under various operating environments.

[0068] The integrated analysis module 6 analyzes the calculated integrated analysis coefficient THX and determines whether an alarm needs to be triggered based on the magnitude of THX. Specifically, when THX ≤ 1, it indicates that the current battery health is good and no alarm is needed; when THX > 1, it indicates that the battery has a potential fault risk or communication instability, requiring an alarm to be triggered. This automatic alarm mechanism based on comprehensive evaluation avoids the false alarms and missed alarms that may occur with traditional single alarm systems. By combining multiple battery health indicators, the system can accurately determine when an alarm needs to be triggered, greatly improving the reliability and intelligence of the battery management system. Especially in underwater environments, the system can respond in real time and issue accurate alarms, reducing human intervention and judgment errors, and improving operational safety.

[0069] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0070] The specific embodiments of the invention have been described in detail above, but they are only examples, and this application is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this application. Therefore, all equivalent changes, modifications, and improvements made without departing from the spirit and principles of this application should be covered within the scope of this application.

Claims

1. An intelligent equalization underwater battery health monitoring and maintenance system, characterized in that: It includes a multimodal sensing module (1), an edge computing processing module (2), a hybrid equalization module (3), a fault prediction and protection module (4), and an adaptive communication module (5). The multimodal sensing module (1) is used to collect multiple data, preprocess and dimensionless the multiple data, and reorganize the processed data into a first data group, a second data group and a third data group. The edge computing processing module (2) is used to perform data coupling calculations; The hybrid equalization module (3) is used to cooperate with the edge computing processing module (2) to perform coupled calculations on the first data group, the second data group and the third data group, thereby obtaining the mode adjustment coefficient, and analyzing the mode adjustment coefficient to perform mode adjustment; The fault prediction and protection module (4) is used to cooperate with the edge computing processing module (2) to perform coupled calculations on the first data group, the second data group and the third data group, and then obtain the fault warning coefficient for analysis, so as to determine whether there is a fault risk in the current mode; The adaptive communication module (5) is used to cooperate with the edge computing processing module (2) to perform data coupling calculation on the first data group, the second data group and the third data group, thereby obtaining the communication detection coefficient, and analyzing the communication detection coefficient to determine whether the communication is stable in the current mode; The integrated analysis module (6) is used to couple all parameters under the first data group, the second data group and the third data group to generate integrated analysis coefficients, and analyze them. Based on the analysis results, it is determined whether the optimized underwater battery needs to be directly alarmed. The feedback module (7) is used to feed back various parameters and analysis results to the visualization terminal.

2. The intelligent, equalized, underwater battery health monitoring and maintenance system of claim 1, wherein, The multimodal sensing module (1) includes a multi-parameter sensing unit (11) and a data fusion processing unit (12). The multi-parameter sensing unit (11) is used to acquire multiple parameters of the underwater battery by deploying multi-parameter sensors and fiber optic sensing technology, including voltage fluctuation coefficient, current harmonic distortion rate, impedance phase angle, coulomb efficiency, temperature non-uniformity, thermal time constant, heat dissipation efficiency, thermal runaway risk index, capacity decay slope, internal resistance growth rate, polarization voltage increment and aging consistency index. The data fusion processing unit (12) performs noise extraction on the acquired raw parameters, performs spatiotemporal alignment of the data, performs feature extraction and data compression, and reorganizes them into a first data group, a second data group and a third data group. The first data set includes voltage fluctuation coefficient A, current harmonic distortion rate B, impedance phase angle C, and coulomb efficiency D. The second set of data includes temperature non-uniformity E, thermal time constant F, heat dissipation efficiency G, and thermal runaway risk index H. The third data set includes the capacity decay slope I, the internal resistance growth rate J, the polarization voltage increment K, and the aging consistency index L.

3. The intelligent, equalized, underwater battery health monitoring and maintenance system of claim 2, wherein, The hybrid equalization module (3) includes an active equalization control unit (31) and an intelligent switching management unit (32). The active equalization control unit (31) is used to extract parameters from the first data group, the second data group and the third data group, and input them into the edge computing processing module (2). The edge computing processing module (2) performs feature fusion through a multi-layer neural network to calculate and obtain the mode adjustment coefficient. The intelligent switching management unit (32) is used to perform data analysis on the calculated mode adjustment coefficients, and determine whether mode switching is required based on the analysis results. The specific analysis method is as follows: when When this time, it means that the underwater battery does not need to switch modes. when When this occurs, it indicates that the underwater battery needs to switch modes.

4. The intelligent equalization underwater battery health monitoring and maintenance system according to claim 3, characterized in that, The edge computing processing module (2) calculates and obtains the mode adjustment coefficient using the following formula; In the formula: A is the voltage fluctuation coefficient, E is the temperature non-uniformity, J is the internal resistance growth rate, and L is the aging consistency index.

5. The intelligent equalization underwater battery health monitoring and maintenance system according to claim 4, characterized in that, The fault prediction and protection module (4) includes an intelligent diagnosis prediction unit (41) and a hierarchical protection execution unit (42). The intelligent diagnostic prediction unit (41) is used to extract parameters from the first data group, the second data group and the third data group, including, and input them into the edge computing processing module (2). The edge computing processing module (2) performs feature fusion through a multi-layer neural network to calculate and obtain the fault warning coefficient. The graded protection execution unit (42) is used to perform data analysis on the calculated fault warning coefficients, and to determine whether there is a fault risk in the current mode based on the analysis results. The specific steps are as follows: when This indicates that there is currently no risk of failure for the underwater battery; when This indicates that the current underwater battery is at risk of failure.

6. The intelligent equalization underwater battery health monitoring and maintenance system according to claim 5, characterized in that, The edge computing processing module (2) calculates and obtains the fault warning coefficient using the following formula; In the formula: B is the current harmonic distortion rate, E is the temperature non-uniformity, H is the thermal runaway risk index, I is the capacity decay slope, J is the internal resistance growth rate, K is the polarization voltage increment, and L is the aging consistency index.

7. The intelligent equalization underwater battery health monitoring and maintenance system according to claim 6, characterized in that, The adaptive communication module (5) includes a multi-mode communication control unit (51) and a data optimization transmission unit (52). The multi-mode communication control unit (51) is used to extract the first data group, the second data group and the third data group, including inputting the extracted parameters into the edge computing processing module (2), and performing feature fusion through the multi-layer neural network of the edge computing processing module (2) to calculate and obtain the communication detection coefficient; The data optimization and transmission unit (52) is used to analyze the calculated data and determine whether the communication is stable under the current mode based on the analysis results. The specific steps are as follows: when This indicates that the underwater battery signal transmission is currently unstable. when At this time, it indicates that the underwater battery needs to send a stable signal.

8. The intelligent equalization underwater battery health monitoring and maintenance system according to claim 7, characterized in that, The edge computing processing module (2) calculates the communication detection coefficient using the following formula: In the formula: A is the voltage fluctuation coefficient, B is the current harmonic distortion rate, C is the impedance phase angle, D is the coulomb efficiency, E is the temperature non-uniformity, F is the thermal time constant, G is the heat dissipation efficiency, H is the thermal runaway risk index, K is the polarization voltage increment, I is the capacity decay slope, and L is the aging consistency index.

9. The intelligent equalization underwater battery health monitoring and maintenance system according to claim 8, characterized in that, The integrated analysis module (6) inputs the parameters from the first data group, the second data group and the third data group into the edge computing processing module (2), and performs feature fusion through a multi-layer neural network by the edge computing processing module (2) to calculate and obtain the integrated analysis coefficients. In the formula: TJX is the mode adjustment coefficient, GZY is the fault warning coefficient, and CDC is the communication detection coefficient; The integrated analysis module (6) analyzes the calculated data and, based on the analysis results, determines whether the optimized underwater battery needs to trigger a direct alarm. The specific method is as follows: when When this time, it means that the underwater battery does not need to trigger a direct alarm. when When this occurs, it indicates that the underwater battery needs to issue an immediate alarm.

Images