An artificial intelligence-based energy management system for electric marine vessels

Abstract

The application discloses an energy management system for an electric ship based on artificial intelligence, and relates to the technical field of energy management. The system collects battery data in real time, calculates a health state coefficient to determine the state of the battery, and if the state is normal, distributes power consumption according to the remaining power and the health state coefficient, and synchronously monitors the performance degradation trend. If the health state is abnormal or a continuous accelerated degradation is monitored, a warning is triggered, realizing intelligent management and early risk intervention. The application combines internal resistance and temperature data, realizes accurate monitoring of early battery degradation through a health state quantitative model and a dynamic threshold, intelligently allocates load in combination with the state of charge and the health state, optimizes the working condition of the battery pack and prolongs the service life, analyzes the change trend of the health state coefficient to warn the accelerated performance degradation, constructs a real-time closed-loop warning between the ship and the shore, improves the collaborative maintenance capability, and comprehensively guarantees the energy safety and efficient operation of the electric ship.

Description

Technical Field

[0001] This invention relates to the field of energy management technology, specifically to an artificial intelligence-based energy management system for electric ships. Background Technology

[0002] In the harsh operating environment of electric ships, the reliability and energy efficiency of the battery system are crucial to navigation safety. Existing technologies cannot accurately predict battery performance degradation or identify degradation trends in advance, leading to delayed problem handling and the accumulation of hidden dangers. Especially in ocean voyages with limited support conditions, this passive management mode not only directly threatens navigation safety but also seriously restricts the improvement of the ship's overall energy efficiency. Summary of the Invention

[0003] Technical problems to be solved To address the shortcomings of existing technologies, this invention provides an artificial intelligence-based electric ship energy management system, which solves the problem that existing technologies cannot predict battery performance degradation in a timely manner under the harsh operating environment of ships, resulting in passive management that threatens navigation safety and restricts the improvement of energy efficiency.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an artificial intelligence-based electric ship energy management system, comprising the following specific modules: a data acquisition module: real-time acquisition of the remaining power of each battery pack, the total power demand of the electric ship, the actual internal resistance of each battery pack, and the temperature data of each battery pack; a health anomaly analysis module: based on the actual internal resistance and temperature data of each battery pack, a comprehensive calculation is performed to obtain the health status coefficient of each battery pack, and the health status of the battery pack is analyzed accordingly. If the analysis indicates that the health status of the battery pack is normal, the battery pack power allocation module is executed; if the analysis indicates that the health status of the battery pack is abnormal... If the battery pack's status is abnormal, the health anomaly warning module is executed; the battery pack power allocation module calculates the current allocation weight based on the remaining power of each battery pack and its health status coefficient, allocates the required power to each battery pack according to the current allocation weight, and simultaneously executes the performance degradation analysis module; the performance degradation analysis module identifies and counts the time points of accelerated performance degradation of the battery pack and the number of affected units, and then analyzes whether the performance of this battery pack is continuously accelerating its decline. If the analysis shows continuous accelerated decline, the health anomaly warning module is also executed; if the analysis shows no continuous accelerated decline, the process returns to the data acquisition module or ends directly; the health anomaly warning module provides warnings.

[0005] Furthermore, the specific method for obtaining the health status coefficient of each battery pack is as follows: based on the factory specifications of the battery pack, the initial internal resistance and the upper limit of the safe temperature of the battery pack are preset. The ratio of the initial internal resistance of each battery pack to the actual internal resistance of the battery pack is calculated to obtain the safe status value of each battery pack. The difference between the upper limit of the safe temperature of each battery pack and the temperature data value of the battery pack is calculated to obtain the fault tolerance temperature value of each battery pack. The fault tolerance temperature value and the safe status value of each battery pack are combined and normalized to obtain the health status coefficient of each battery pack. ;in, This represents the health status coefficient of each battery pack. This indicates the safety status value of each battery pack. This indicates the tolerance temperature value for each battery pack.

[0006] Furthermore, the specific method for analyzing the health status of the battery pack in the health anomaly analysis module is as follows: a preset health status threshold for the battery pack is used, and the health status coefficient of each battery pack is compared with the health status threshold. If the health status coefficient of the battery pack is greater than the health status threshold, it indicates that the health status of the battery pack is normal. If the health status coefficient of the battery pack is less than or equal to the health status threshold, it indicates that the health status of the battery pack is abnormal.

[0007] Furthermore, the specific method for obtaining the current allocation weight is as follows: Based on the factory specifications of the battery pack, the rated total capacity of the battery pack is preset. The ratio of the remaining capacity of each battery pack to the rated total capacity of the battery pack is calculated to obtain the state of charge coefficient of each battery pack. The current allocation weight is obtained by multiplying the state of charge coefficient of each battery pack with the health state coefficient of each battery pack and then normalizing the result.

[0008] Furthermore, the specific method for identifying and statistically analyzing the time points and the number of affected battery packs with accelerated performance degradation is as follows: A coordinate system for the change of the health status coefficient of each battery pack is established, where the horizontal axis of the coordinate system represents the time series, and the vertical axis represents the health status coefficient of the battery pack. Several coordinate points are then obtained and recorded as the health coordinate points of each battery pack. Based on the health coordinate points of each battery pack, the trend of continuous accelerated performance degradation of the battery pack is calculated, and battery packs with continuously accelerating performance degradation are recorded as abnormal battery packs. The number of abnormal battery packs is counted to obtain an abnormal quantity value. The duration of continuous accelerated performance degradation of the abnormal battery packs is counted to obtain an abnormal time value.

[0009] Furthermore, the specific method for calculating the continuous accelerated decline trend of the performance of each battery pack based on the health coordinate points of each battery pack is as follows: calculate the slope between the previous health coordinate point and the next health coordinate point in turn to obtain several slope values; analyze whether the performance of the battery pack is continuously declining based on the several slope values; when the performance of the battery pack is continuously declining, calculate the absolute value of the difference between the previous slope value and the next slope value in turn to obtain several acceleration values; analyze whether the performance of the battery pack is continuously accelerating its decline based on the several acceleration values.

[0010] Furthermore, the specific method for analyzing whether the performance of the battery pack continues to decline based on several slope values ​​is as follows: compare any slope value of the battery pack with zero in turn. If all slope values ​​are less than zero, it indicates that the performance of the battery pack continues to decline. If there is any slope value greater than or equal to zero, it indicates that the performance of the battery pack does not continue to decline.

[0011] Furthermore, the specific method for analyzing whether the performance of the battery pack continues to accelerate its decline based on several acceleration values ​​is as follows: the subsequent acceleration value is compared with the previous acceleration value in turn. If the subsequent acceleration value is greater than the previous acceleration value, it indicates that the performance of the battery pack continues to accelerate its decline. If the subsequent acceleration value is less than or equal to the previous acceleration value, it indicates that the performance of the battery pack does not continue to accelerate its decline.

[0012] Furthermore, the specific method for analyzing whether the performance of this battery pack continues to accelerate its decline is as follows: analyze whether the abnormal quantity value is equal to the number of battery packs installed on the ship, and analyze whether the abnormal time value exceeds the preset detection time value. If the abnormal quantity value is equal to the number of battery packs installed on the ship or the abnormal time value does not exceed the preset detection time value, it indicates that the performance of the battery pack does not continue to accelerate its decline. If the abnormal quantity value is not equal to the number of battery packs installed on the ship and the abnormal time value exceeds the preset detection time value, it indicates that the performance of the battery pack continues to accelerate its decline.

[0013] Beneficial effects

[0014] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: 1. By comprehensively calculating the ratio of the actual internal resistance of the battery to the initial internal resistance, the system obtains the safe state value, the difference between the upper limit of the safe temperature and the actual temperature, and the fault tolerance temperature value. Furthermore, it introduces a health state coefficient calculation formula based on an exponential function, which can amplify and sensitively respond to minute changes such as battery internal resistance degradation and temperature rise risk. Combined with the use of Min-Max normalization processing and a dynamic threshold setting method that integrates the initial threshold and unsupervised learning, the system achieves accurate and adaptive evaluation of the health state of each battery pack, overcoming the shortcomings of traditional methods that are insensitive to early performance degradation.

[0015] 2. When allocating power, the system not only considers the remaining charge of the battery, i.e., the state of charge coefficient, but also innovatively incorporates its health state coefficient. The current allocation weight is obtained by multiplying the two and normalizing them. Based on this, the power demand is allocated to each battery pack. This allocation strategy can automatically tilt the load towards battery packs with high state of charge and good health while ensuring energy supply, thereby optimizing the overall working state of the battery pack, extending the battery system life, and improving the energy efficiency and safety of the ship during operation.

[0016] 3. By establishing a coordinate system for the health status coefficient to change over time, calculating the slope between consecutive time points and the absolute value of the difference between slopes, and analyzing the abnormal quantity and abnormal time values, the system can effectively distinguish between the overall performance degradation caused by common factors and the irreversible accelerated degradation of individual batteries. This mechanism can issue an early warning when the battery health status coefficient has not fallen below the static threshold but has shown a continuous and accelerated downward trend, thus realizing early and proactive identification of faults.

[0017] 4. When the system determines that the battery health is abnormal or its performance is deteriorating rapidly, the health abnormality early warning module will generate a structured alarm containing information such as battery number, internal resistance, and temperature. This alarm not only triggers an audible and visual alarm in the ship's bridge via Ethernet, but also automatically sends it to the remote monitoring platform in a standard message format via shipborne satellite communication or mobile network. At the same time, the system fully records the abnormal time sequence data and logs. This design ensures that potential risks can be perceived in real time by both local crew members and remote maintenance personnel, supporting rapid decision-making and maintenance, and forming a zero-delay ship-shore collaborative safety closed loop.

[0018] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0019] Figure 1 This invention relates to a flowchart of an artificial intelligence-based energy management method for electric ships.

[0020] Figure 2 This invention relates to a structural diagram of an artificial intelligence-based electric ship energy management system.

[0021] Figure 3 This invention relates to a trend coordinate graph of the health status of a battery pack. Detailed Implementation

[0022] 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.

[0023] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0024] Example 1:

[0025] like Figures 1-2 As shown, this embodiment of the invention provides an artificial intelligence-based electric ship energy management system, which includes the following specific modules: Data acquisition module: The battery pack of electric ships adopts a modular design. Its configuration and layout can be flexibly adjusted according to the tonnage and range requirements of the ship. For example, small ships usually install one or several battery packs in the cabin, while large ships, in order to meet the energy and power requirements of megawatt-hours, will distribute dozens or even hundreds of standardized battery packs in the form of battery compartments similar to containers at the bottom or side of the hull. All battery packs are managed by a central BMS battery management system and fixed by a marine-grade structure. They are equipped with a tight liquid cooling or air cooling system and fireproof compartments to ensure safe and stable operation in complex maritime environments. The BMS (Battery Management System) achieves precise control through multi-dimensional collaborative monitoring: It collects filtered and denoised voltage and current time-series data in real time, and uses a Kalman filter algorithm to dynamically estimate the remaining power of each battery pack, providing a core basis for predicting ship range and energy scheduling decisions. Simultaneously, it obtains real-time total power demand from the EMS (Energy Management System) via the ship's Ethernet. The EMS collects and aggregates power consumption data from all shipboard electrical equipment, such as propulsion motors, auxiliary machinery, and living facilities, to accurately calculate the ship's total power demand at any given moment, using this as the overall target for battery pack power output. Furthermore, it calculates the actual internal resistance of each battery pack by injecting current pulses and measuring the voltage transient response, thereby assessing battery health and providing early warnings of potential failure risks. Finally, it relies on a network of temperature sensors distributed across the battery pack surface to directly collect temperature data from each battery pack, constructing a safety guarantee against thermal runaway.

[0026] Health Anomaly Analysis Module: Based on the factory specifications of the battery packs, the module presets the initial internal resistance and upper limit of the safe temperature for each battery pack. It calculates the ratio of the initial internal resistance to the actual internal resistance of each battery pack to obtain the safe state value for each pack. It then calculates the difference between the upper limit of the safe temperature and the actual temperature data for each battery pack to obtain the fault tolerance temperature value. Finally, it performs a comprehensive calculation of the fault tolerance temperature and the safe state value for each battery pack, followed by normalization. Specifically, based on historical operating data, it determines the minimum and maximum values ​​for both the fault tolerance temperature and the safe state value for each battery pack. Then, using the Min-Max normalization method, it calculates the fault tolerance temperature for each battery pack. The normalized values ​​of the temperature tolerance and the normalized values ​​of the safety status of each battery pack are calculated. The normalized value of the fault tolerance temperature of each battery pack is equal to the fault tolerance temperature of each battery pack minus its historical minimum value and divided by the difference between the historical maximum and minimum values. The normalized value of the safety status of each battery pack is equal to the safety status of each battery pack minus its historical minimum value and divided by the difference between the historical maximum and minimum values. Finally, the normalized fault tolerance temperature and the normalized safety status of each battery pack are substituted into the formula for calculating the health status coefficient of each battery pack. This eliminates the difference in dimensions and realizes the comprehensive calculation of the fault tolerance temperature and the safety status of each battery pack, thus obtaining the health status coefficient of each battery pack. The system presets health status thresholds for battery packs. Initial fixed thresholds are set based on battery factory specifications and aging test data. Their core function is to provide a reliable safety baseline for the system in the early, data-scarce phase, and serve as the physical safety lower limit that subsequent dynamic thresholds must not exceed. After the system accumulates data, unsupervised learning algorithms, such as Gaussian mixture models, are used to cluster the health status coefficients of each historical battery pack to identify the state distribution of the battery group. Then, using the initial fixed thresholds as a reference, warning thresholds that better reflect the current actual health status of the batteries are dynamically generated based on the cluster centers and boundaries. Finally, the entire threshold system is iteratively optimized by continuously comparing learning results and adding new abnormal cases. The health status coefficient of each battery pack is compared with the health status threshold of the battery pack. If the health status coefficient of the battery pack is greater than the health status threshold of the battery pack, it means that the health status of the battery pack is normal, and then the battery pack power allocation module is executed. If the health status coefficient of the battery pack is less than or equal to the health status threshold of the battery pack, it means that the health status of the battery pack is abnormal, and then the health abnormality warning module is executed.

[0027] Battery pack power allocation module: Based on the factory specifications of the battery pack, the rated total capacity of the battery pack is preset. The remaining capacity of each battery pack is calculated as a ratio to its rated total capacity to obtain the state of charge (SOC) coefficient of each battery pack. The higher the remaining capacity of a battery pack, the higher its SOC coefficient. The SOC coefficient of each battery pack is then multiplied by its state of health coefficient and normalized using a Min-Max normalization method to obtain the current allocation weight. Based on this weight, the required power consumption is allocated to each battery pack. ,in, Indicates the first The power consumption required by the battery pack Indicates the first Current allocation weights for battery packs This represents the total electricity demand, and =1, where The number of battery packs installed on the ship is determined to ensure that the total power demand is fully and accurately distributed to each battery pack, and the performance degradation analysis module is executed synchronously.

[0028] Performance degradation analysis module: Identifies and statistically analyzes the time points and number of affected battery packs that experience accelerated performance degradation, and then analyzes whether the battery pack performance continues to degrade at an accelerated rate, thus providing early warning of faults. If the analysis shows that the battery pack performance is continuously degrading at an accelerated rate, even if the battery pack's health status is normal, the accelerated performance degradation indicates that the battery pack's health status is declining at an accelerated rate, and the health anomaly warning module is also executed. If the analysis shows that the battery pack performance is not continuously degrading at an accelerated rate, the process returns to the data acquisition module or terminates directly.

[0029] Health Anomaly Early Warning Module: First, it generates structured early warning information including the battery pack number, actual internal resistance, and temperature data of the battery pack indicating an abnormal health status. Subsequently, this information is uploaded in real time to the ship's bridge via the electric vessel's Ethernet for audible and visual alarms, and simultaneously automatically sent to the ship manufacturer's remote monitoring platform in a standard alarm message format via shipborne satellite communication or mobile network. At the same time, the system records complete anomaly time series data and early warning logs for subsequent diagnosis and maintenance decisions, thereby ensuring a zero-delay closed-loop response to potential risks from the ship to the shore.

[0030] Example 2 differs from Example 1 in that: The specific method for obtaining the health status coefficient of each battery pack is as follows: ; in, This represents the health status coefficient of each battery pack, reflecting the health status of each battery pack. This indicates the safety status value of each battery pack, reflecting its overall safety condition. A lower safety status value indicates weaker performance of the battery pack. This represents the fault tolerance temperature value of each battery pack, reflecting the fault tolerance temperature of each battery pack. The smaller the fault tolerance temperature value of a battery pack, the higher the risk of that battery pack. The health status coefficient of each battery pack is in exponential form, which can amplify small changes in the safety status value and the fault tolerance temperature value of each battery pack. This amplification effect enables the BMS to provide sensitive early warning of potential safety degradation and temperature anomalies in the battery pack. In addition, the larger the safety status value or the larger the fault tolerance temperature value of the battery pack, the healthier the battery pack is.

[0031] The specific methods for identifying and statistically analyzing the time points of accelerated performance degradation of battery packs and the number of affected units are as follows: A coordinate system for the health status coefficient changes of each battery pack is established through a subroutine embedded in the BMS battery management system. The horizontal axis of the coordinate system represents the time series, and the vertical axis represents the health status coefficient of the battery pack. This yields several coordinate points, denoted as the health coordinate points for each battery pack. Figure 3As shown, the vertical axis represents the health status coefficient calculated based on an exponential function, indicating the health level of the battery pack; the horizontal axis represents the time series in minutes, representing the continuous monitoring period. When the trend of the coordinate points shows a continuous decline, that is, when the slope between all adjacent points is negative, it indicates that the performance of the battery pack is continuously declining. When the rate of decline of the coordinate points is constantly accelerating, that is, the difference between adjacent slopes is monotonically increasing, a warning of continuous accelerated degradation is triggered. This figure reveals how the system can achieve forward-looking identification and risk intervention of continuous accelerated battery degradation by tracking the change trajectory of the health status coefficient. Based on the health coordinate points of each battery pack, the continuous accelerated decline trend of the battery pack's performance is calculated, and battery packs with continuously accelerating performance decline are recorded as abnormal battery packs. The number of abnormal battery packs is counted to obtain the abnormal quantity value, and the continuous accelerated performance decline time of abnormal battery packs is counted to obtain the abnormal time value.

[0032] The specific method for calculating the continuously accelerating decline in the performance of each battery pack based on its health coordinate points is as follows: The slope between the previous and next healthy coordinate points is calculated sequentially to obtain several slope values. Based on these slope values, it is analyzed whether the performance of the battery pack continues to decline. If the performance of the battery pack continues to decline, the absolute value of the difference between the previous and next slope values ​​is calculated sequentially to obtain several acceleration values. Based on these acceleration values, it is analyzed whether the performance of the battery pack continues to decline at an accelerated rate.

[0033] The specific method for analyzing whether the performance of this battery pack is continuously declining based on several slope values ​​is as follows: Since the slope mathematically represents the trend of change, if the slope of the health status coefficient with time, i.e. the first derivative, is continuously negative, it means that the health status coefficient of this battery pack is monotonically decreasing with time, which directly corresponds to the physical fact of the continuous degradation of battery performance. Therefore, any slope value of this battery pack is compared with zero in turn. If the slope value is less than zero, it means that the performance of this battery pack is continuously declining. If there is any slope value greater than or equal to zero, it means that the performance of this battery pack is not continuously declining.

[0034] The specific method for analyzing whether the performance of this battery pack is continuously and rapidly declining based on several acceleration values ​​is as follows: The subsequent acceleration value is compared with the previous acceleration value in turn. If the subsequent acceleration value is greater than the previous acceleration value, it means that the performance of this battery pack is continuously accelerating its decline. That is, the acceleration value is monotonically increasing, which means that the acceleration of the decline is continuously increasing. The corresponding performance of this battery pack is not only fast, but also constantly getting faster. If the subsequent acceleration value is less than or equal to the previous acceleration value, it means that the performance of this battery pack is not continuously accelerating its decline.

[0035] The specific method for analyzing whether the performance of this battery pack continues to degrade at an accelerated pace is as follows: The analysis determines whether the abnormal quantity value equals the number of battery packs installed on the ship, and whether the abnormal time value exceeds a preset detection time value. This detection time value is initially determined by conducting accelerated aging tests on the battery packs in the laboratory and then using high-frequency data statistical analysis. Subsequently, it is deployed and verified under typical operating conditions of electric ships. By evaluating the correlation between warning signals and actual degradation at different time windows and combining operational feedback for iterative optimization, a practical time value is finally obtained that can effectively capture degradation trends and avoid false alarms in real navigation environments. If the analysis shows that the abnormal quantity value equals the number of battery packs installed on the ship or the abnormal time value does not exceed a preset detection time value, then the detection time value is considered valid. If the preset detection time value indicates that the battery pack performance has not continued to accelerate its decline, that is, if all battery packs are judged to be abnormal, it means that all battery packs are affected by the same external factor, such as harsh operating conditions, cooling failure, or improper operation, resulting in a synchronous decline in performance. Or, if the abnormal time value does not exceed the preset detection time value, it means that the accelerated performance decline of this battery pack is reversible. That is, if the accelerated decline is caused by sudden changes in operating conditions, instantaneous temperature rise, or measurement noise, it does not mean that irreversible degradation has occurred inside the battery pack. If the analysis shows that the number of abnormal values ​​is not equal to the number of battery packs installed on the ship and the abnormal time value exceeds the preset detection time value, it means that the battery pack performance continues to accelerate its decline.

[0036] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. An artificial intelligence-based energy management system for electric ships, characterized in that: Includes the following specific modules: Data acquisition module: Real-time acquisition of the remaining power of each battery pack, the total power demand of the electric ship, the actual internal resistance of each battery pack, and the temperature data of each battery pack; Health anomaly analysis module: Based on the actual internal resistance and temperature data of each battery pack, a comprehensive calculation is performed to obtain the health status coefficient of each battery pack, and the health status of the battery pack is analyzed accordingly. If the analysis shows that the health status of the battery pack is normal, the battery pack power allocation module is executed. If the analysis indicates that the health status of this battery pack is abnormal, then the health abnormality warning module will be executed. Battery pack power allocation module: Based on the remaining power of each battery pack and the health status coefficient of each battery pack, a comprehensive calculation is performed to obtain the current allocation weight. The required power consumption is allocated to each battery pack according to the current allocation weight, and the performance degradation analysis module is executed simultaneously. Performance degradation analysis module: Identifies and counts the time points and the number of affected battery packs that accelerate performance degradation, and then analyzes whether the performance of this battery pack continues to accelerate. If the analysis shows that the performance is continuously accelerating, the health anomaly warning module is also executed; if the analysis shows that the performance is not continuously accelerating, the data acquisition module is returned or the process ends directly. The specific method for identifying and statistically analyzing the time points of accelerated performance degradation of battery packs and the number of affected units is as follows: Establish a coordinate system for the change of health status coefficients for each battery pack, where the horizontal axis of the coordinate system is the time series and the vertical axis is the health status coefficient of the battery pack. This yields several coordinate points, which are denoted as the health coordinate points for each battery pack. Based on the health coordinate points of each battery pack, calculate the continuous accelerating decline trend of its performance. Battery packs with continuously accelerating performance decline are denoted as abnormal battery packs. Count the number of abnormal battery packs to obtain the abnormal quantity value. Count the duration of continuous accelerating performance decline for abnormal battery packs to obtain the abnormal time value. The specific method for calculating the continuously accelerating decline in the performance of a battery pack based on the health coordinate points of each battery pack is as follows: The slope between the previous and next healthy coordinate points is calculated sequentially to obtain several slope values. Based on these slope values, it is analyzed whether the performance of the battery pack continues to decline. If the performance of the battery pack continues to decline, the absolute value of the difference between the previous and next slope values ​​is calculated sequentially to obtain several acceleration values. Based on these acceleration values, it is analyzed whether the performance of the battery pack continues to decline at an accelerated rate. The specific method for analyzing whether the performance of this battery pack is continuously declining based on several slope values ​​is as follows: Compare any slope value of this battery pack with zero in turn. If all slope values ​​are less than zero, it means that the performance of this battery pack is continuously declining. If there is any slope value greater than or equal to zero, it means that the performance of this battery pack is not continuously declining. The specific method for analyzing whether the performance of this battery pack continues to decline at an accelerating rate based on several acceleration values ​​is as follows: The subsequent acceleration value is compared with the previous acceleration value in turn. If the subsequent acceleration value is greater than the previous acceleration value, it means that the performance of this battery pack is continuously accelerating down. If the subsequent acceleration value is less than or equal to the previous acceleration value, it means that the performance of this battery pack is not continuously accelerating down. The specific method for analyzing whether the performance of this battery pack continues to degrade at an accelerated rate is as follows: The system analyzes whether the abnormal quantity value is equal to the number of battery packs installed on the ship and whether the abnormal time value exceeds the preset detection time value. If the abnormal quantity value is equal to the number of battery packs installed on the ship or the abnormal time value does not exceed the preset detection time value, it indicates that the battery pack performance has not continued to accelerate its decline. If the abnormal quantity value is not equal to the number of battery packs installed on the ship and the abnormal time value exceeds the preset detection time value, it indicates that the battery pack performance has continued to accelerate its decline. Health Anomaly Early Warning Module: Provides early warnings.

2. The artificial intelligence-based electric ship energy management system according to claim 1, characterized in that: The specific method for obtaining the health status coefficient of each battery pack is as follows: Based on the factory specifications of the battery pack, the initial internal resistance and the upper limit of the safe temperature of the battery pack are preset. The ratio of the initial internal resistance of each battery pack to the actual internal resistance of the battery pack is calculated to obtain the safe state value of each battery pack. The difference between the upper limit of the safe temperature of each battery pack and the temperature data value of the battery pack is calculated to obtain the fault tolerance temperature value of each battery pack. The fault tolerance temperature value and the safe state value of each battery pack are combined and normalized to obtain the health state coefficient of each battery pack. ； in, This represents the health status coefficient of each battery pack. This indicates the safety status value of each battery pack. This indicates the tolerance temperature value for each battery pack.

3. The artificial intelligence-based electric ship energy management system according to claim 2, characterized in that: The specific method by which the health status analysis module analyzes the battery pack's health status is as follows: A preset health status threshold for the battery pack is established. The health status coefficient of each battery pack is compared with the health status threshold. If the health status coefficient of a battery pack is greater than the health status threshold, it indicates that the health status of the battery pack is normal. If the health status coefficient of a battery pack is less than or equal to the health status threshold, it indicates that the health status of the battery pack is abnormal.

4. The artificial intelligence-based electric ship energy management system according to claim 3, characterized in that: The specific method for obtaining the current allocation weights is as follows: Based on the factory specifications of the battery pack, the rated total capacity of the battery pack is preset. The ratio of the remaining capacity of each battery pack to the rated total capacity of the battery pack is calculated to obtain the state of charge coefficient of each battery pack. The current distribution weight is obtained by multiplying the state of charge coefficient of each battery pack with the health state coefficient of each battery pack and then normalizing the result.

Images