Beacon light battery performance prediction method

Abstract

The invention discloses a beacon light battery performance prediction method, and relates to the field of beacon light battery performance prediction, and the method comprises the steps: constructing a battery multi-dimensional original feature data set containing time sequence information, and generating an electricity storage performance feature set and a charging and discharging speed feature; calculating an electricity storage performance index and a charging and discharging performance index, and further fusing to generate a battery health index; by calculating an electricity storage performance index and a charging and discharging performance index and further fusing the indexes to generate a battery health index, a set of health degree quantitative evaluation system based on the internal characteristics of the battery is established. By deeply fusing the internal electrochemical state and the external performance of the battery, the interpretability of the performance evaluation process is realized, the actual health state of the battery can be accurately reflected, the future performance degradation trend can be effectively predicted, a scientific basis is provided for operation and maintenance of the beacon light battery, and the normal operation of the beacon light is prevented from being influenced by sudden failure of the battery.

Description

Technical Field

[0001] This invention belongs to the field of navigation light battery performance prediction technology, and relates to a method for predicting the performance of navigation light batteries. Background Technology

[0002] Navigation lights are critical safety facilities for maritime navigation, and their stable operation directly affects the safety of ship passage. Currently, navigation lights generally use batteries as their primary power source, and the degradation of battery performance directly affects their lighting duration and operational reliability. Therefore, accurate assessment and prediction of the performance status of navigation light batteries is of significant practical importance.

[0003] Currently, existing technologies have proposed methods for predicting the performance of navigation light batteries. For example, invention patent CN117633717A proposes a method and system for predicting the performance of navigation light batteries. This method collects time series data of navigation light battery operating parameters and environmental information. Through parameter correlation analysis and random forest parameter importance assessment, it extracts key operating parameter sequences. These key parameter sequences, environmental information, and historical performance time series are then input into two performance prediction models. The overall battery performance prediction conclusion is obtained by combining the results of both models. This method can estimate the remaining battery life, providing a basis for timely battery replacement and mitigating the impact of battery performance degradation on the normal use of navigation lights.

[0004] However, while the aforementioned existing technologies provide a multi-model fusion-based solution for predicting the performance of navigation light batteries, they still have the following shortcomings: First, battery performance degradation is directly related to the phased characteristics of the charging and discharging process. Batteries exhibit unique characteristics at different stages of charging and discharging, and these phased characteristics are the core windows reflecting changes in the battery's internal state. The existing technology only inputs the extracted sequence of key battery operating parameters into the prediction model as a whole, without physically deconstructing the phased characteristics of the charging and discharging process. Essentially, this is a black-box approach, making it impossible to accurately locate performance anomalies at each stage of charging and discharging, thus affecting the accuracy of subsequent battery health status assessments.

[0005] Secondly, existing technologies employ multiple prediction models for result fusion, but the collaborative mechanism between these models is not clearly defined. Furthermore, these models assume linearity and stationarity of time series data, while battery performance degradation actually exhibits non-stationary and non-linear trends. Forcing a stationary approach may distort the actual degradation patterns. Additionally, when using convolutional neural networks for time series prediction, extra structures are often required to accommodate temporal dependencies; otherwise, the model may lack the ability to capture long-term time series patterns. With limited data or an imbalanced sample distribution, such models are prone to overfitting or underfitting, thus affecting the stability of the predictions.

[0006] Finally, existing technologies rely solely on trained static models for processing, which cannot adapt to the degradation trend of battery performance over time or the impact of sudden environmental events. This makes the prediction results insensitive to real-time performance fluctuations and difficult to adapt to the actual degradation trend of beacon lights batteries under complex operating conditions. Summary of the Invention

[0007] In view of this, in order to solve the problems mentioned in the background art, the present invention provides a method for predicting the performance of navigation light batteries.

[0008] The objective of this invention can be achieved through the following technical solution: a method for predicting the performance of a navigation light battery, comprising: S1, real-time acquisition of voltage, current and corresponding timestamp data of the navigation light battery during operation, and construction of a multi-dimensional original feature dataset of the battery containing time-series information. S2. Process the original multi-dimensional feature dataset of the battery to generate a set of energy storage performance features that includes the battery's actual energy storage capacity, energy consumption under use, and self-discharge under idle state. Based on the energy storage performance feature set, perform fusion processing to obtain energy storage performance indicators.

[0009] S3. Extract the current-time relationship during the charging and discharging stages respectively to obtain the battery charging speed characteristics and discharging speed characteristics. Calculate the charging and discharging performance indicators based on the charging speed characteristics and discharging speed characteristics.

[0010] S4. The battery health index is obtained based on the energy storage performance index and the charge and discharge performance index.

[0011] S5. Determine the current battery performance level based on the battery health index, and store the level in the historical battery performance level sequence in chronological order.

[0012] S6. Predict the degradation trend of battery performance level within a specified future time period based on historical battery performance level sequence.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention constructs a multi-dimensional original feature dataset of the battery containing time sequence information, and generates a set of energy storage performance features and charging and discharging speed features, transforming the abstract charging and discharging process into quantifiable stage features, so that the performance changes of each stage of battery charging and discharging can be captured, realizing detailed tracking of the internal state of the battery, and more accurately reflecting the actual health state of the battery.

[0014] (2) By calculating energy storage performance indicators and charge / discharge performance indicators, and further integrating them to generate a battery health index, this invention establishes a quantitative evaluation system for battery health based on the intrinsic characteristics of the battery. This effectively overcomes the shortcomings of simply relying on data-driven models, which may distort the actual nonlinear degradation law of the battery, and improves the accuracy of performance evaluation and adaptability to complex working conditions.

[0015] (3) This invention dynamically judges and records the performance level based on the battery health index to form a historical battery performance level sequence, and predicts the future degradation trend based on the change frequency of the sequence, so that the prediction results can respond to the sudden changes in battery performance more timely and make a more stable judgment on the long-term degradation trend, thereby enhancing the robustness and practicality of the prediction model. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a diagram illustrating the implementation steps of the method of the present invention.

[0018] Figure 2 This is a flowchart for determining the battery performance level of the present invention.

[0019] Figure 3 This is a flowchart for predicting future trends of the battery in this invention. Detailed Implementation

[0020] 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.

[0021] Please see Figure 1 As shown, the present invention provides a method for predicting the performance of a navigation light battery, including: S1, real-time acquisition of voltage, current and corresponding timestamp data during the operation of the navigation light battery, and construction of a multi-dimensional original feature dataset of the battery containing time-series information.

[0022] Specifically, voltage and current sensors connected to the beacon light battery are used to collect voltage and current values ​​during battery operation at a fixed frequency in real time. Simultaneously, a real-time clock module generates timestamp data accurate to the millisecond level. Subsequently, the collected raw data is cleaned to remove outliers and a filtering algorithm is used to smooth noise. This results in the construction of a structured time-series dataset arranged in chronological order and containing timestamp, voltage, and current fields. This multi-dimensional raw feature dataset of the battery provides basic data support for subsequent performance analysis.

[0023] S2. Process the original multi-dimensional feature dataset of the battery to generate a set of energy storage performance features that includes the battery's actual energy storage capacity, energy consumption under use, and self-discharge under idle state. Based on the energy storage performance feature set, perform fusion processing to obtain energy storage performance indicators.

[0024] The steps for generating the above energy storage performance feature set are as follows: the interval where the current is positive and monotonically increases until the current returns to zero is defined as the charging interval, and the interval where the current is negative and monotonically decreases until the current returns to zero is defined as the discharging interval.

[0025] It should be noted that when an external charging device supplies electrical energy to the battery, current flows from the charging device to the battery, and the current value corresponding to this process is defined as a positive value. When the battery supplies electrical energy to the beacon light load, current flows from the battery to the load, and the current value corresponding to this process is defined as a negative value.

[0026] During the charging phase, external charging devices typically inject electrical energy into the battery with a gradually increasing current. As the battery charge approaches saturation, the charging current gradually decreases until it reaches zero, completing the charging process and stopping energy input. During the discharging phase, when the battery supplies power to the navigation light load, it initially outputs a relatively stable negative current. As the load operates or the battery charge is consumed, the output current gradually decreases until the battery stops supplying power, and the discharging process ends when the current reaches zero.

[0027] The actual energy storage capacity of the battery is obtained by integrating the current value over time within the charging range.

[0028] Set a current threshold within the charging and discharging range. If the absolute value of the current is greater than the threshold, it is determined to be in use. If the absolute value of the current is less than or equal to the threshold, it is determined to be in idle state.

[0029] It should be noted that the current threshold setting varies depending on the type of navigation light battery, such as lead-acid batteries and lithium batteries, and covers the battery's maximum idle current.

[0030] When the battery is in use, such as when the navigation light load is working normally or the charging equipment is actively charging, energy flows between the battery and the external device at a relatively high intensity, and the corresponding absolute value of the current is relatively large. When the battery is idle, such as when the navigation light is not working or the charging equipment stops charging, the energy flow almost stops, and there is only a slight self-discharge or standby loss, and the corresponding absolute value of the current is extremely small.

[0031] The power consumption during the usage period is obtained by integrating the current value over time.

[0032] The difference between the actual battery charge at the start and end of each idle state is recorded as the self-discharge amount in the idle state.

[0033] The actual energy storage capacity, the energy consumption under use, and the self-discharge under idle conditions are summarized into a set of energy storage performance characteristics.

[0034] The calculation steps for the above energy storage performance indicators are as follows: extract the actual energy storage capacity, energy consumption under use, and self-discharge under idle state from the energy storage performance feature set.

[0035] The rated storage capacity of the battery is determined based on the battery model, and the capacity retention rate is obtained by calculating the ratio of the actual storage capacity to the rated storage capacity.

[0036] The self-discharge rate is calculated based on the total idle time and the cumulative self-discharge amount in the idle state.

[0037] The power consumption rate is calculated based on the total battery usage time and the cumulative power consumption under various usage conditions, using the power consumption per unit time as the power consumption rate.

[0038] Weighting coefficients are assigned to capacity retention rate, self-discharge rate and power consumption rate respectively, and weighted summation is performed to obtain energy storage performance indicators.

[0039] The formula for calculating energy storage performance indicators is as follows: ; in, Indicates energy storage performance indicators, Indicates the actual amount of electricity stored. Indicates the rated storage capacity. Indicates the self-discharge rate. Indicates power consumption rate. This represents the weighting coefficient, used to adjust the degree of influence on performance indicators.

[0040] The specific explanation of the above formula is as follows: This indicates the battery's capacity retention rate, reflecting the ratio of the battery's current energy storage capacity to its rated capacity, and assessing the degree of capacity degradation. The larger the value, the greater the positive impact of capacity retention rate on energy storage performance indicators. As a positive contribution term in the formula, the larger its value, the higher the P-value.

[0041] A higher value indicates a stronger weakening effect of self-discharge rate on energy storage performance indicators. As a negative contribution term in the formula, the larger its value, the lower the P-value.

[0042] The larger the value, the stronger the weakening effect of the power consumption rate on the final performance index of energy storage. As a negative contribution term in the formula, the larger its value, the lower the P-value.

[0043] in The acquisition method is as follows: First, collect the voltage and current time series data of the beacon light battery during a complete charge and discharge cycle and its corresponding performance degradation records to construct a feature-performance correlation dataset. Second, calculate the correlation strength or contribution of the battery's actual energy storage capacity, self-discharge rate, power consumption rate and overall battery health status. Use regression analysis and other methods to quantify the influence weight of each feature on performance evaluation and normalize it so that the sum of the weight coefficients is 1. Finally, obtain the weight coefficient combination that can be used for fusion calculation.

[0044] S3. Extract the current-time relationship during the charging and discharging stages respectively to obtain the battery charging speed characteristics and discharging speed characteristics. Calculate the charging and discharging performance indicators based on the charging speed characteristics and discharging speed characteristics.

[0045] The steps for obtaining the battery charging speed features are as follows: Determine a complete charging stage from the battery's multi-dimensional original feature dataset based on the charging interval.

[0046] During the charging phase, the entire charging phase is divided into multiple charging sub-intervals arranged in chronological order based on the change of current value over time.

[0047] Specifically, during the charging phase, the timing current data is read sequentially. When the current values ​​of multiple consecutive data points are detected to remain at the same level, it is determined that a constant current charging sub-interval has been entered, and its start and end points are marked.

[0048] When the current values ​​of multiple consecutive data points show a monotonically increasing or monotonically decreasing trend, it is determined that a converter charging sub-range has been entered, and its start and end points are marked.

[0049] For each charging sub-interval, the average charging current value and its corresponding charging duration are extracted, and the ratio of the average charging current value to the charging duration is calculated as the unit time charging rate of each sub-interval.

[0050] Arrange the unit time charging rates of all charging sub-intervals in chronological order to form a charging speed characteristic sequence.

[0051] The steps for obtaining the above discharge rate characteristics are as follows: Monitor the changes in battery voltage while the battery is in use. When the voltage values ​​of multiple consecutive data points are detected to remain at the same level, it is identified as a discharge plateau.

[0052] It should be noted that the core of battery discharge is the process of releasing electrical energy through oxidation-reduction reactions of internal active materials. During this process, when the reaction rate of the active materials and the internal ion migration rate are in a relatively stable equilibrium state, the battery terminal voltage will remain within a relatively fixed range without significant fluctuations, thus creating a discharge plateau.

[0053] For each discharge platform, calculate the average absolute value of its discharge current, record the duration of the platform, and calculate the discharge rate per unit time.

[0054] The calculation method is as follows: divide the product of the average discharge current of the platform and the duration by the duration of the platform to obtain the discharge rate per unit time.

[0055] All discharge platforms are arranged in chronological order of their occurrence, and the corresponding discharge rate per unit time is extracted to form a discharge velocity feature sequence.

[0056] The steps for obtaining the above charging and discharging performance indicators are as follows: construct a charging speed versus time curve from the charging speed feature sequence, calculate the long-term trend slope of the curve, if the slope is greater than or equal to zero, set the charging severity coefficient to zero, if the slope is negative and its absolute value exceeds the preset trend threshold, calculate the charging severity coefficient based on the absolute value of the long-term trend slope.

[0057] The long-term trend slope of the charging speed versus time curve was calculated using linear regression fitting.

[0058] It should be noted that the preset trend threshold is determined based on the model characteristics of the navigation light battery, historical charging data, and performance degradation standards.

[0059] When the long-term trend slope of the curve is greater than or equal to 0, it indicates that there is no degradation in charging performance, so the coefficient is set to zero. When the slope is negative and the absolute value exceeds the preset trend threshold, it indicates that the degradation of charging performance has exceeded the normal fluctuation range. The larger the absolute value of the slope, the faster the charging speed drops and the more serious the degradation of charging performance. The coefficient value is also larger accordingly, providing core parameters that reflect the degree of degradation of charging performance for the subsequent calculation of charging and discharging performance indicators and battery health index.

[0060] The charging severity coefficient is obtained by dividing the long-term trend slope of the charging speed characteristic sequence by a preset charging slope degradation benchmark. This benchmark value, based on battery type and empirically set parameters, represents the minimum degradation slope that warrants attention. The calculation result is then normalized.

[0061] The discharge rate characteristic sequence is used to construct a discharge rate versus time curve. The long-term trend slope of the curve is calculated. If the long-term trend slope of the discharge rate characteristic sequence is greater than or equal to zero and the long-term trend slope of the discharge plateau duration is zero, then the discharge severity coefficient is set to zero.

[0062] The long-term trend slope of the discharge rate versus time curve was calculated using linear regression fitting.

[0063] It should be noted that if the slope of the long-term trend of the discharge rate versus time curve is greater than or equal to 0, it means that the discharge efficiency has not decreased over time. This implies that the reaction capacity of the active materials inside the battery and the ion migration efficiency have not deteriorated. If there are no obvious problems such as electrode aging or increased internal resistance, the core indicators of discharge performance are in a normal state.

[0064] During long-term use, a healthy battery should maintain a stable discharge platform duration under the same type of load, with a long-term trend slope equal to 0. This indicates that the battery's energy supply continuity during the stable discharge phase has not decreased, further confirming that there has been no substantial deterioration in discharge performance.

[0065] Otherwise, the discharge severity coefficient is calculated based on the absolute value of the long-term trend slope of the discharge rate versus time curve and the trend slope of the discharge plateau duration.

[0066] It should be noted that, since discharge performance and stability jointly determine discharge health, the discharge severity coefficient is obtained by normalizing the weighted fusion of the absolute value of the long-term trend slope of the discharge rate versus time curve and the trend slope of the discharge plateau duration.

[0067] Based on preset basic performance values, charging severity coefficient, and discharging severity coefficient, the charging and discharging performance index values ​​are calculated.

[0068] The formulas for calculating the above charge / discharge performance indicators are as follows: ; in, This indicates the charging and discharging performance index, reflecting the overall level of battery charging and discharging performance; the higher the value, the better the performance. This represents the baseline performance value, which is a benchmark value for charge and discharge performance set based on the battery model or historical data. This represents the severity coefficient of charging, reflecting the degree of degradation in charging performance. Its value ranges from 0 to 1, where 0 indicates no degradation and 1 indicates complete degradation. The discharge severity coefficient reflects the degree of degradation of discharge performance. The value ranges from 0 to 1, where 0 indicates no degradation and 1 indicates complete degradation.

[0069] The specific explanation above is as follows: This item reflects the impact of charging performance on the overall index. When the charging severity coefficient is 0, the value of this item is 1, indicating that there is no degradation in charging performance and no negative impact on the overall index. When the charging severity coefficient increases, the value of this item decreases, reducing the overall index and reflecting the impact of deteriorating charging performance.

[0070] Similarly, This item reflects the impact of discharge performance on the overall index. When the discharge severity coefficient is 0, the value of this item is 1, indicating that there is no deterioration in discharge performance and no negative impact on the overall index. When the discharge severity coefficient increases, the value of this item decreases, reducing the overall index and reflecting the impact of discharge performance deterioration.

[0071] It is emphasized that performance degradation of either party will lead to a decrease in the index, and the greater the degree of degradation, the more significant the decrease in the index.

[0072] S4. The battery health index is obtained based on the energy storage performance index and the charge and discharge performance index.

[0073] The steps to obtain the battery health index are as follows: obtain the energy storage performance index and the charge / discharge performance index.

[0074] If the energy storage performance index is less than zero, the energy storage performance index will be adjusted to zero; if the energy storage performance index is greater than or equal to zero, the energy storage performance index will remain unchanged.

[0075] It should be noted that energy storage performance indicators essentially reflect the quality of a battery's energy storage capacity. This capacity is at its lowest when the battery completely loses its energy storage function, corresponding to an indicator value of zero. Negative energy storage capacity is impossible. For example, the minimum capacity retention rate is 0, and the weighted calculation result of self-discharge rate and power consumption rate will not cause the overall indicator to exceed the lower limit of the capacity. If extreme abnormal data, such as calculation deviations caused by data acquisition errors, results in a negative indicator value, this negative value does not conform to the actual boundary of battery energy storage performance and needs to be adjusted to zero to restore the physical meaning of the indicator.

[0076] The battery health index is obtained by multiplying the adjusted energy storage performance index by the charge / discharge performance index.

[0077] It should be noted that the multiplication operation emphasizes the interdependence between energy storage performance and charge / discharge performance. This non-linear combination method highlights the coupling effect between performance indicators, so that poor performance of any one indicator will have a significant negative impact on the health index, thus more intuitively reflecting the overall health status of the battery.

[0078] S5. Determine the current battery performance level based on the battery health index, and store the level in the historical battery performance level sequence in chronological order.

[0079] See Figure 2 As shown, the specific steps for determining the current battery performance level are as follows: extract the battery health index in real time and store it in the historical battery health index sequence.

[0080] If the amount of data in the historical battery health index sequence does not meet the preset minimum data requirement, the current battery performance level is set to the initial performance level; otherwise, the most recent battery health index is extracted from the historical battery health index sequence as the reference health index.

[0081] It should be noted that the minimum data requirement must ensure that the reference value is used when the most recent index is referenced. For example, historical test data verification shows that when the number of health index sequence data for the same type of navigation light battery is greater than or equal to 5, the comparison error using the most recent data as a reference is less than 5%; if the number of data is less than 5, the error will exceed 10%. Therefore, based on this verification result, the minimum data number can be set to 5 to ensure that the error of the reference comparison is within an acceptable range.

[0082] If the historical health index sequence data volume does not reach the preset minimum data volume, it indicates that there is insufficient historical data available for comparison. The difference between the current and historical data cannot accurately reflect the true changes in battery performance. Forcibly calculating the difference to determine the level in this case would be inaccurate. Using the initial level as the current level avoids confusion caused by missing data and provides an initial benchmark for subsequent difference comparisons after accumulating sufficient historical data, ensuring the continuity of performance level evaluation.

[0083] When the amount of historical data reaches the standard, the most recent health index best reflects the battery's performance status in the previous stage. By comparing it with the current index, the recent performance change trend can be accurately captured, ensuring the timeliness and accuracy of the level judgment.

[0084] Calculate the difference between the current battery health index and the reference health index. If the difference is positive, the current battery performance level will be upgraded by one performance level based on the latest performance level.

[0085] It should be noted that a positive difference between the current battery health index and the reference health index indicates that the battery's core performance, such as energy storage capacity and charge / discharge efficiency, has improved compared to the previous stage. For example, after maintenance, the battery's internal resistance decreases and its energy storage capacity is restored. This positive change in performance needs to be reflected through an upgrade in performance level to ensure that the level is consistent with the actual health status.

[0086] If the difference is negative, the current battery performance level will be reduced by one performance level from the latest performance level.

[0087] It should be noted that when the difference between the current health index and the reference health index is negative, it indicates that the core performance of the current battery, such as its energy storage capacity and charge / discharge efficiency, has deteriorated compared to the previous stage. This could be due to factors such as electrode aging leading to increased internal resistance, decreased energy storage capacity, or slower charge / discharge speed. Such negative performance changes need to be reflected by a reduction in the performance level to ensure that the performance status directly corresponds to the level label and to avoid assessment bias where performance has deteriorated but the level has not been updated accordingly.

[0088] The steps for determining the performance level order are as follows: Obtain the historical battery performance level sequence. If there are no records of performance level changes in the sequence, then based on the initial performance state of the battery, a performance level list is preset in descending order of performance level.

[0089] It should be noted that, based on the battery's initial performance state, such as its health state at the time of manufacture, a list of performance levels from best to worst is preset, corresponding to the optimal or higher initial level. This allows for the direct assignment of a clear initial level to the current battery, avoiding situations where the performance level cannot be determined due to a lack of change records, and ensuring that the performance level assessment has a unified standard throughout the battery's entire life cycle.

[0090] Conversely, if the battery health index shows a continuous decline before the recording point and then tends to stabilize after the recording point, the recording point is determined to be an effective performance inflection point.

[0091] It should be noted that initially, due to issues such as electrode aging and increased internal resistance, the health index will continuously decline. Once this decline reaches a certain level, performance will stabilize, and the health index will no longer decrease. The continuous decline before the recording point indicates that performance is in the process of degradation, while the stabilization after the recording point indicates the end of degradation. This characteristic of decline followed by stabilization perfectly matches the true transition of battery performance from dynamic degradation to static stability. Determining this recording point as an effective performance inflection point allows for accurate identification of the core change node in battery performance status, avoiding misjudging occasional fluctuations as inflection points.

[0092] For inflection points where performance level increases after a recording point, move the corresponding new level forward in the priority order list; for inflection points where performance level decreases after a recording point, move the corresponding new level backward in the priority order list.

[0093] It should be noted that the performance ranking list, from best to worst, indicates that the earlier a performance level is, the better the performance, and the later a performance level is, the worse the performance. When a performance level improvement occurs after a recording point, the new level outperforms the previous level, and the new level should be placed at a higher position to match its actual performance. Conversely, when a performance level decline occurs, the new level is worse than the previous level, and moving its position to the back ensures that the new level is placed at a lower position to reflect its performance.

[0094] If the difference is zero, then the current battery performance level is the same as the latest performance level.

[0095] It should be noted that a zero difference between the current battery health index and the reference health index means that the current health index is completely consistent with the reference health index, indicating that the battery's current energy storage capacity, charge and discharge efficiency, and other core performance characteristics have not changed.

[0096] S6. Predict the degradation trend of battery performance level within a specified future time period based on historical battery performance level sequence.

[0097] See Figure 3 As shown, the specific content of the above prediction of the degradation trend of battery performance level within a specified future time period is as follows: count the number of times the performance level changes in the historical battery performance level sequence within the most recent specified time period, and calculate its change frequency.

[0098] If the frequency of change is higher than the preset frequency threshold, the battery performance is determined to be in an unstable state, and it is predicted that the performance level will continue to degrade within a specified period of time in the future.

[0099] It should be noted that the preset frequency threshold is determined based on battery model characteristics, historical healthy battery data, navigation light usage scenarios, and performance degradation risk standards.

[0100] A healthy and stable battery will not experience frequent fluctuations in its core performance, such as energy storage capacity and charge / discharge efficiency. The frequency of changes in its corresponding performance level will remain at a low level. However, when the battery performance is unstable, the core performance will change frequently, directly leading to frequent rises and falls in performance level, which manifests as an increased frequency of change.

[0101] Conversely, if the battery performance is stable, it is predicted that the performance level will remain at the current level for a specified period of time in the future, and the degradation trend will stop.

[0102] It should be noted that when the core performance of the battery is in a stable state, such as stable internal electrode reaction, no significant fluctuation in internal resistance, and good electrolyte condition, its health index will not change frequently, thus the performance level does not need to be adjusted frequently, and the frequency of change is lower than or equal to the preset threshold.

[0103] Predicting the degradation trend of battery performance levels over a specified period can help navigation light batteries plan maintenance in advance and avoid affecting the normal operation of navigation lights due to sudden battery performance failures.

[0104] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0105] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0106] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0107] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0108] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for predicting the performance of navigation light batteries, characterized in that: include: S1. Real-time acquisition of voltage, current and corresponding timestamp data of the navigation light battery during operation, and construction of a multi-dimensional original feature dataset of the battery containing time-series information; S2. Process the original feature dataset of the battery in multiple dimensions to generate a set of energy storage performance features that includes the actual energy storage capacity, energy consumption under use and self-discharge under idle state. Based on the energy storage performance feature set, perform fusion processing to obtain the energy storage performance index. S3. Extract the current-time relationship during the charging and discharging stages respectively to obtain the battery charging speed characteristics and discharging speed characteristics. Calculate the charging and discharging performance indicators based on the charging speed characteristics and discharging speed characteristics. S4. Obtain the battery health index based on energy storage performance indicators and charge / discharge performance indicators; S5. Determine the current battery performance level based on the battery health index and store the level in the historical battery performance level sequence in chronological order. S6. Predict the degradation trend of battery performance level within a specified future time period based on historical battery performance level sequence.

2. The method for predicting the performance of a navigation light battery according to claim 1, characterized in that: The steps for generating the energy storage performance feature set are as follows: The interval where the current is positive and monotonically increases until the current returns to zero is defined as the charging interval, and the interval where the current is negative and monotonically decreases until the current returns to zero is defined as the discharging interval. The actual stored capacity of the battery is obtained by integrating the current value over time within the charging range. Set a current threshold within the charging and discharging range. If the absolute value of the current is greater than the threshold, it is determined to be in use. If the absolute value of the current is less than or equal to the threshold, it is determined to be in idle state. The power consumption during the usage period is obtained by integrating the current value over time. The difference between the actual battery charge at the start and end of each idle state is recorded as the self-discharge amount in the idle state. The actual energy storage capacity, the energy consumption under use, and the self-discharge under idle conditions are summarized into a set of energy storage performance characteristics.

3. The method for predicting the performance of a navigation light battery according to claim 1, characterized in that: The calculation steps for the energy storage performance indicators are as follows: The actual energy storage capacity, energy consumption under usage, and self-discharge under idle state are extracted from the energy storage performance characteristics set. The rated storage capacity of the battery is determined based on the battery model, and the capacity retention rate is obtained by calculating the ratio of the actual storage capacity to the rated storage capacity. The self-discharge rate is calculated based on the total idle time and the cumulative self-discharge amount in the idle state. The power consumption rate is calculated based on the total battery usage time and the cumulative power consumption under usage conditions, using the power consumption per unit time as the power consumption rate. Weighting coefficients are assigned to capacity retention rate, self-discharge rate and power consumption rate respectively, and weighted summation is performed to obtain energy storage performance indicators.

4. The method for predicting the performance of a navigation light battery according to claim 1, characterized in that: The steps for obtaining the battery charging speed characteristics are as follows: A complete charging stage is determined from the battery's multi-dimensional raw feature dataset based on the charging interval; During the charging phase, the entire charging phase is divided into multiple charging sub-intervals arranged in chronological order based on the change of current value over time. For each charging sub-interval, extract its average charging current value and its corresponding charging duration, and calculate the ratio of the average charging current value to the charging duration as the unit time charging rate of each sub-interval. Arrange the unit time charging rates of all charging sub-intervals in chronological order to form a charging speed characteristic sequence.

5. The method for predicting the performance of a navigation light battery according to claim 1, characterized in that: The steps for obtaining the discharge velocity characteristics are as follows: The system monitors changes in battery voltage during battery use. When multiple consecutive data points show that the voltage values ​​remain at the same level, it is identified as a discharge plateau. For each discharge platform, calculate the average absolute value of its discharge current, record the duration of the platform, and calculate the discharge rate per unit time. All discharge platforms are arranged in chronological order of their occurrence, and the corresponding discharge rate per unit time is extracted to form a discharge velocity feature sequence.

6. The method for predicting the performance of a navigation light battery according to claim 1, characterized in that: The steps for obtaining the charge / discharge performance indicators are as follows: The charging speed feature sequence is used to construct a charging speed versus time curve. The long-term trend slope of the curve is calculated. If the slope is greater than or equal to zero, the charging severity coefficient is set to zero. If the slope is negative and its absolute value exceeds the preset trend threshold, the charging severity coefficient is calculated based on the absolute value of the long-term trend slope. The discharge rate characteristic sequence is used to construct a discharge rate versus time curve. The long-term trend slope of the curve is calculated. If the long-term trend slope of the discharge rate characteristic sequence is greater than or equal to zero and the long-term trend slope of the duration of the discharge plateau is zero, then the discharge severity coefficient is set to zero. Otherwise, the discharge severity coefficient is calculated based on the absolute value of the long-term trend slope of the discharge rate versus time curve and the trend slope of the discharge plateau duration. Based on preset basic performance values, charging severity coefficient, and discharging severity coefficient, the charging and discharging performance index values ​​are calculated.

7. The method for predicting the performance of a navigation light battery according to claim 1, characterized in that: The steps for obtaining the battery health index are as follows: Obtain energy storage performance indicators and charge / discharge performance indicators; If the energy storage performance index is less than zero, the energy storage performance index will be adjusted to zero; if the energy storage performance index is greater than or equal to zero, the energy storage performance index will remain unchanged. The battery health index is obtained by multiplying the adjusted energy storage performance index by the charge / discharge performance index.

8. The method for predicting the performance of a navigation light battery according to claim 1, characterized in that: The specific steps for determining the current battery performance level based on the battery health index are as follows: Real-time extraction and storage of battery health index into historical battery health index sequence; If the amount of data in the historical battery health index sequence does not meet the preset minimum data requirement, the current battery performance level is set to the initial performance level; otherwise, the most recent battery health index is extracted from the historical battery health index sequence as the reference health index. Calculate the difference between the current battery health index and the reference health index. If the difference is positive, the current battery performance level will be upgraded by one performance level based on the latest performance level. If the difference is negative, the current battery performance level will be reduced by one performance level from the latest performance level. If the difference is zero, then the current battery performance level is the same as the latest performance level.

9. The method for predicting the performance of a navigation light battery according to claim 8, characterized in that: The steps for determining the performance level order are as follows: Obtain the historical battery performance level sequence. If there are no records of performance level changes in the sequence, then based on the battery's initial performance state, pre-determine a performance level list from best to worst. Conversely, when the battery health index shows a continuous decline before the recording point and then tends to stabilize after the recording point, the recording point is determined to be an effective performance inflection point. For inflection points where performance level increases after a recording point, move the corresponding new level forward in the priority order list; for inflection points where performance level decreases after a recording point, move the corresponding new level backward in the priority order list.

10. The method for predicting the performance of a navigation light battery according to claim 9, characterized in that: The specific details of predicting the degradation trend of battery performance level within a specified future time period are as follows: Count the number of times the performance level changes within the most recent specified time period in the historical battery performance level sequence, and calculate its change frequency; If the frequency of change is higher than the preset frequency threshold, the battery performance is determined to be in an unstable state, and it is predicted that the performance level will continue to degrade within a specified period of time in the future. Conversely, if the battery performance is stable, it is predicted that the performance level will remain at the current level for a specified period of time in the future, and the degradation trend will stop.

Images