Portable power supply module life prediction and early warning method

Abstract

The invention provides a portable power supply module life prediction and early warning method, and belongs to the module early warning technical field, and the method comprises the following steps: collecting multi-dimensional monitoring data of a power supply module through an integrated sensor group, generating an operation simulation scene linked with an actual operation scene in real time based on the data fitting, and dynamically mapping the real-time working condition and the spatial position of the power supply module. And constructing a digital model matched with the actual specification in the simulation scene, judging the aging process of the core component, and calculating the residual maintenance life of the power supply module. Graded early warning is triggered according to the remaining maintenance life and the data abnormal degree, maintenance prompts are pushed through local alarm and remote dual-mode communication, full-cycle data are synchronously recorded, and support is provided for fault tracing and model optimization. The portable power supply module life prediction and early warning method provided by the invention realizes preventive maintenance, adapts to communication conditions of different field environments, is high in automation degree, does not need frequent manual intervention, and reduces the maintenance cost of field operation.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of module early warning, more specifically, to a portable power module life prediction and early warning method. BACKGROUND

[0002] The portable power module is the core power supply equipment for field operation, and its operation stability directly determines the task execution efficiency and safety. In the scenes of field support of the army, outdoor engineering construction, etc., the power module often faces harsh environments such as high and low temperature, sand, vibration, humidity change, etc., and needs to withstand frequent charge and discharge cycles, which leads to rapid aging of core components and high risk of sudden failure.

[0003] The existing maintenance method of the portable power module mainly relies on artificial regular inspection, which has obvious limitations: the inspection cycle is fixed, the aging state of the module cannot be grasped in real time, the fault prediction is lagging behind, and the problem can only be found after the module fails, which leads to interruption of field operation; the inspection only relies on manual observation of appearance and measurement of basic electrical parameters, it is difficult to obtain the real aging data of the internal core components, and the judgment accuracy is poor; the existing few life prediction methods only focus on single electrical parameters such as charge and discharge times, without considering the acceleration effect of field environmental factors on aging, the prediction accuracy is insufficient, and it cannot adapt to complex field scenes; the early warning mechanism is single, only triggers an alarm after a fault occurs, lacks graded early warning and preventive maintenance guidance, and is difficult to meet the high requirements of power supply continuity in tasks such as field support of the army. SUMMARY

[0004] The purpose of the present application is to provide a portable power module life prediction and early warning method, which aims to solve the problem of more areas needing to be processed on the floating disc, more components needing to be installed, inconvenient maintenance, and low integration level.

[0005] To achieve the above purpose, the technical scheme adopted by the present application is: a portable power module life prediction and early warning method is provided, comprising: Collecting multi-dimensional monitoring data of the power module by an integrated sensor group, the monitoring data at least covering the operating state, field environment and physical dynamic characteristic data of the power module; Generating an operating simulation scene in real time linkage with the actual operation scene based on the monitoring data, the operating simulation scene dynamically mapping the real-time working condition and spatial position of the power module; Building a digital model matched with the actual power module specifications in the operating simulation scene, the digital model having a built-in core component aging evolution law database; The monitoring data is input into the digital model in real time, the acceleration influence of the environment and dynamic characteristics on the aging of the core components is combined to comprehensively determine the aging process and calculate the remaining maintenance life of the power module; according to the remaining maintenance life and the data anomaly degree, a hierarchical early warning is triggered, a maintenance prompt is pushed through local alarm and remote dual-mode communication, and full-cycle data is recorded for fault tracing and model optimization.

[0006] In a possible implementation, the multi-dimensional monitoring data of the power module collected by the integrated sensor group includes: The running state includes the number of charge-discharge cycles, output voltage stability, internal core component temperature, charge-discharge rate, external input power fluctuation and idle storage time length of the power module, which are collected by an adaptive sensor unit; The field environment includes the temperature and humidity and vibration intensity of the working area, which are collected by sensors deployed at specific positions of the module; the characteristic data includes the appearance integrity, surface damage condition and dynamic parameters in the moving process of the key parts of the power module, which are collected by corresponding sensor units.

[0007] In a possible implementation, the running simulation scene in real-time linkage with the actual working scene is generated based on the monitoring data, including: The monitoring data is preprocessed to eliminate invalid data, eliminate transient interference and standardize different dimension data; a sensor regular self-calibration mechanism is set, and the acquisition parameters are corrected by a built-in standard reference source at a preset period to avoid data drift; The scene modeling algorithm is used to fuse the preprocessed data and geographic spatial information to construct a dynamically updated running simulation scene, ensuring consistency with the actual scene.

[0008] In a possible implementation, the digital model has a built-in core component aging evolution law database, including: The database is constructed based on a large number of actual application cases, contains aging period data of battery cells, circuit boards, heat dissipation components and the like under different conditions, and presets multi-dimensional aging acceleration coefficients; the digital model and the running simulation scene establish a real-time data interaction link to ensure that the parameters of the digital model are dynamically matched with the actual working conditions of the power module; the digital model has data learning ability and can iteratively optimize the associated data and acceleration coefficients in the core component aging evolution law database based on historical prediction bias.

[0009] In a possible implementation, the hierarchical early warning triggered according to the remaining maintenance life and the data anomaly degree includes: The digital model matches real-time data with historical data in the database, calls the corresponding acceleration coefficient to calculate the aging percentage of core components, weights the overall aging degree, and estimates the remaining maintenance life. The graded early warning mechanism includes at least three levels of early warning, with different early warning levels corresponding to different remaining maintenance life ranges and data anomalies. When triggered, maintenance guidance is pushed. The remote communication supports automatic switching of multiple modes to adapt to field conditions.

[0010] In one possible implementation, the simultaneous recording of full-cycle data for fault tracing and model optimization includes: The built-in storage module records the monitoring data, early warning records, and maintenance operation information for each charge and discharge cycle, and sets a data retention period. Maintenance personnel can read the data from the built-in storage module through a dedicated terminal for fault root cause analysis and maintenance process optimization. The digital model automatically stores prediction process data and regularly updates the core component aging evolution law database to improve prediction accuracy.

[0011] In one possible implementation, the step of generating a real-time operational simulation scenario based on the monitoring data includes: Based on the scenario modeling algorithm, geospatial data of the work area is imported to accurately map the real-time location of the power module and associate it with surrounding environmental interference sources. Based on the physical dynamic feature data, the appearance of the power module is visualized in the operation simulation scenario, and potential damage areas are marked. The movement trajectory and stress of the power module are simulated by combining dynamic data. Abnormal data triggers the priority update of the operation simulation scenario, and abnormal states are reported first.

[0012] In one possible implementation, triggering the tiered early warning based on the remaining maintenance life and the degree of data anomaly includes: Import the geographic information and preset action route data of the power module for subsequent operations; reproduce the subsequent geographic environment and action route in the operation simulation scenario, and predict the aging rate change of the core components in combination with the database; calculate the remaining maintenance life of multiple subsequent time nodes based on the predicted rate, generate a life change trend curve, and trigger a forward warning in advance if the predicted life will enter the warning range.

[0013] In one possible implementation, triggering the tiered early warning based on the remaining maintenance life and the degree of data anomaly includes: When multiple power modules are in the same operating area, the digital models of each module share data through remote communication; based on the shared data, an aging correlation model of the regional power cluster is constructed to analyze the mutual influence of aging among the modules; when any module in the cluster triggers an early warning, the correlation model is used to predict the impact on other modules and push collaborative maintenance suggestions.

[0014] In one possible implementation, the simultaneous recording of full-cycle data for fault tracing and model optimization includes: Based on the time-series characteristics of the monitoring data and changes in scene parameters, an aging causative factor feature library is constructed; by comparing real-time data with the causative factor feature library through machine learning algorithms, the core aging causative factors and their contribution ratios are identified. While outputting the remaining maintenance life, a cause diagnosis report is generated, which provides specific recommendations to slow down aging and optimize subsequent operations.

[0015] The beneficial effects of the portable power module lifespan prediction and early warning method provided by this invention are as follows: Compared with the prior art, this invention's portable power module lifespan prediction and early warning method collects multi-dimensional monitoring data of the power module through an integrated sensor group, covering core information such as operating status, field environment, and physical dynamic characteristics. Based on this data, a real-time operational simulation scenario is generated, dynamically mapping the real-time operating conditions and spatial location of the power module. A digital model matching the actual specifications is constructed in the simulation scenario, which has a built-in database of aging evolution laws of core components. After real-time input of monitoring data, the aging process of core components is determined by comprehensively considering the accelerated impact of the environment and dynamic characteristics on aging, and the remaining maintenance life of the power module is estimated.

[0016] Based on the remaining maintenance life and the degree of data anomalies, tiered early warnings are triggered. Maintenance prompts are pushed through local alarms and remote dual-mode communication, and full-cycle data is recorded simultaneously to support fault tracing and model optimization. The method provided in this application ensures the stability and reliability of data acquisition, provides full-process maintenance guidance from routine inspections to emergency replacements, enables preventative maintenance, and avoids sudden failures. It is adaptable to different communication conditions in field environments, has a high degree of automation, requires no frequent manual intervention, reduces maintenance costs in field operations, and is particularly suitable for high-efficiency operational scenarios such as military field support. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art 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.

[0018] Figure 1 A flowchart of a portable power module lifespan prediction and early warning method provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0020] Please see Figure 1 The present invention will now describe the portable power module lifespan prediction and early warning method provided by the present invention. The portable power module lifespan prediction and early warning method includes: Multi-dimensional monitoring data of the power module is collected through an integrated sensor array. The monitoring data covers at least the operating status of the power module, the field environment, and physical dynamic characteristics.

[0021] Based on the monitoring data fitting, a real-time operation simulation scenario is generated and linked with the actual operation scenario. The operation simulation scenario dynamically maps the real-time operating conditions and spatial location of the power module.

[0022] A digital model matching the specifications of the actual power module is built in the simulation scenario. The digital model has a built-in database of the aging evolution law of core components.

[0023] The monitoring data is input into the digital model in real time. The aging process is comprehensively judged and the remaining maintenance life of the power module is estimated by combining the accelerated impact of environmental and dynamic characteristics on the aging of core components. The tiered early warning is triggered based on the remaining maintenance life and the degree of data anomaly. Maintenance prompts are pushed through local alarms and remote dual-mode communication. At the same time, the full-cycle data is recorded for fault tracing and model optimization.

[0024] The beneficial effects of the portable power module lifespan prediction and early warning method provided by this invention are as follows: Compared with the prior art, this invention's portable power module lifespan prediction and early warning method collects multi-dimensional monitoring data of the power module through an integrated sensor group, covering core information such as operating status, field environment, and physical dynamic characteristics. Based on this data, a real-time operational simulation scenario is generated, dynamically mapping the real-time operating conditions and spatial location of the power module. A digital model matching the actual specifications is constructed in the simulation scenario, which has a built-in database of aging evolution laws of core components. After real-time input of monitoring data, the aging process of core components is determined by comprehensively considering the accelerated impact of the environment and dynamic characteristics on aging, and the remaining maintenance life of the power module is estimated.

[0025] Based on the remaining maintenance life and the degree of data anomalies, tiered early warnings are triggered. Maintenance prompts are pushed through local alarms and remote dual-mode communication, and full-cycle data is recorded simultaneously to support fault tracing and model optimization. The method provided in this application ensures the stability and reliability of data acquisition, provides full-process maintenance guidance from routine inspections to emergency replacements, enables preventative maintenance, and avoids sudden failures. It is adaptable to different communication conditions in field environments, has a high degree of automation, requires no frequent manual intervention, reduces maintenance costs in field operations, and is particularly suitable for high-efficiency operational scenarios such as military field support.

[0026] In some embodiments of the portable power module lifespan prediction and early warning method provided in this application, the collection of multi-dimensional monitoring data of the power module through an integrated sensor array includes: The operating status includes the number of charge / discharge cycles of the power module, output voltage stability, internal core component temperature, charge / discharge rate, external input power fluctuations, and idle storage time, all collected by an adapted sensor unit.

[0027] The field environment data includes temperature, humidity, and vibration intensity of the work area, collected by sensors deployed at specific locations on the module; the feature data includes the appearance integrity, surface damage, and dynamic parameters of key parts of the power module during movement, collected by corresponding sensor units.

[0028] Multi-dimensional monitoring data acquisition focuses on "full-element coverage + field adaptability," clarifying the acquisition logic and sensor design requirements through detailed data types. Operational status data focuses on the electrical characteristics of the power module. A charge / discharge cycle counter installed at the charge / discharge interface records the number of cycles (directly related to cell aging). A voltage sensor monitors voltage fluctuations at a frequency of 1 time / second (reflecting circuit stability). A temperature sensor is embedded in the cell and circuit board surface to collect internal temperature (high temperature is a key factor in aging). Current sensors, input voltage monitoring sensors, and timing modules respectively capture charge / discharge rates, external power fluctuations, and idle time, comprehensively covering electrical aging influencing factors. Field environment data acquisition balances protection and accuracy. Temperature and humidity sensors are installed inside the casing to avoid sand and rain corrosion, and vibration sensors are deployed at the bottom of the module to sense the intensity of bumps, ensuring that the data accurately reflects the impact of outdoor temperature, humidity, and vibration on the module. Physical dynamic characteristic data fills gaps in existing technology. Damage to key components such as the sensor acquisition interface and casing is monitored by the appearance, and motion sensors record parameters such as movement speed and bump frequency, filling blind spots in the module's physical state monitoring.

[0029] All sensor units must meet harsh environment resistance standards, with housings featuring IP67 or higher sealing designs to resist sand and dust, and short-term water immersion. Operating temperatures cover -40℃ to 60℃. In low-temperature environments, compensation algorithms prevent data drift, while in high-temperature environments, heat dissipation structures control component temperatures. Furthermore, sensor deployment locations have been optimized through simulation; temperature sensors are positioned away from heat dissipation holes to prevent false readings, and vibration sensors are kept away from power interfaces to reduce electromagnetic interference. This ensures that all data acquisition errors are controlled within 5%, providing a high-quality data source for subsequent scenario modeling and lifespan prediction.

[0030] In some embodiments of the portable power module lifespan prediction and early warning method provided in this application, the operation simulation scenario generated based on monitoring data and linked in real time with the actual operation scenario includes: The monitoring data is preprocessed to remove invalid data, eliminate transient interference, and standardize data of different dimensions; a sensor self-calibration mechanism is set up to correct the acquisition parameters through the built-in standard reference source at a preset cycle to avoid data drift.

[0031] A scenario modeling algorithm is used to integrate preprocessed data and geospatial information to construct a dynamically updated simulation scenario, ensuring consistency with the actual scenario.

[0032] The construction of a simulated scenario involves a process of "data preprocessing - sensor calibration - scenario modeling," with the primary focus on addressing data quality and scenario realism. In the data preprocessing stage, invalid data such as voltage zeroing and temperatures exceeding 100℃ are removed using logical thresholds. Moving averages are used to smooth continuously fluctuating data such as vibration intensity, eliminating instantaneous interference. Then, a standardization algorithm converts data with different dimensions, such as voltage and temperature, into dimensionless data in the 0-1 range, ensuring data can be fused and calculated during subsequent modeling and avoiding scenario deviations caused by data format differences. The sensor's periodic self-calibration mechanism addresses accuracy degradation over long-term use. Calibration is automatically initiated every 30 charge-discharge cycles or monthly. Using a built-in standard reference source (such as a 3.7V standard voltage source and a 25℃ standard temperature source), the collected values ​​are compared with standard values ​​to calculate the deviation rate. If the deviation exceeds 5%, the collected parameters are automatically adjusted, ensuring that the collected accuracy remains within ±2% within a one-year cycle, resolving prediction errors caused by sensor drift.

[0033] Scene modeling adopts a fusion approach of "geospatial information + multi-dimensional data." It imports geographic data such as terrain, altitude, and environmental interference sources (e.g., high-temperature heat sources, strong electromagnetic fields) of the work area through algorithms like digital twins, linking the module's real-time location to the geographic environment. Simultaneously, pre-processed operational status, environmental, and physical dynamic data are mapped into the scene in real time, forming a dynamically visualized scene updated every 5 seconds. For example, when the module moves to a high-temperature area, the scene synchronously displays the increase in ambient temperature and the corresponding change in internal battery cell temperature, intuitively replicating actual working conditions and providing realistic operational scenario support for the digital model.

[0034] In some embodiments of the portable power module lifespan prediction and early warning method provided in this application, the digital model incorporates a database of core component aging evolution patterns, including: The database is built based on a large number of real-world application cases, including aging cycle data of battery cells, circuit boards, heat dissipation components, etc. under different conditions, and presets multi-dimensional aging acceleration coefficients; the digital model establishes a real-time data interaction link with the operation simulation scenario to ensure that the parameters of the digital model are dynamically matched with the actual operating conditions of the power module; the digital model has data learning capabilities and can iteratively optimize the associated data and acceleration coefficients in the aging evolution law database of core components based on historical prediction deviations.

[0035] The core support of the digital model is an aging evolution law database, which is built based on a massive number of real-world application cases in the field, covering aging data of core components such as battery cells, circuit boards, and heat dissipation components under different conditions. For example, it records key data such as battery cells retaining 80% capacity after 1000 cycles at room temperature (25℃), and only retaining 60% capacity after 1000 cycles at high temperature (45℃); circuit board solder joint lifespan of 2000 hours when vibration acceleration is ≤10m / s², and lifespan reduced to 800 hours when vibration acceleration is ≥20m / s²; and preset multi-dimensional aging acceleration coefficients, such as increasing the aging rate of battery cells by 30%-50% in high-temperature environments and increasing the aging rate of circuit board solder joints by 20%-40% in strong vibration environments, providing a quantitative basis for lifespan calculation.

[0036] A real-time data interaction link is established between the digital model and the operational simulation scenario. Changes in parameters such as ambient temperature and vibration intensity in the scenario are synchronously fed back to the model, adjusting the corresponding aging acceleration coefficient to ensure that the model parameters dynamically match the actual module operating conditions. Simultaneously, the model possesses self-learning capabilities, periodically comparing historical predicted lifespans with actual aging results and calculating the deviation. If the deviation exceeds 8%, the aging cycle data and acceleration coefficient in the database are iteratively optimized. For example, if the aging rate of a circuit board in a high-humidity environment is 15% faster than recorded in the database, the acceleration coefficient for that environment is updated, continuously improving the model's prediction accuracy.

[0037] In some embodiments of the portable power module lifespan prediction and early warning method provided in this application, triggering graded early warnings based on remaining maintenance lifespan and the degree of data anomaly includes: The digital model matches real-time data with historical data in the database, calls the corresponding acceleration coefficient to calculate the aging percentage of core components, weights the overall aging degree and estimates the remaining maintenance life; the graded early warning mechanism includes at least three levels of early warning, with different early warning levels corresponding to different remaining maintenance life ranges and data anomalies, and pushes maintenance guidance when triggered, and remote communication supports automatic switching of multiple modes to adapt to field conditions.

[0038] The remaining service life estimation relies on the multi-parameter fusion calculation logic of the digital model. The model first performs similarity matching between real-time multi-dimensional data (such as the current number of charge / discharge cycles, ambient temperature, and vibration intensity) and historical data in the aging evolution law database, selecting the closest combination of operating scenarios and calling the corresponding aging acceleration coefficient. Then, it calculates the aging percentage of each core component separately. For example, the battery cell is estimated to be 70% aged based on the number of cycles and temperature, and the circuit board is estimated to be 50% aged based on vibration intensity. Subsequently, the overall aging degree of the module is calculated by weighting the components according to their functional weights (such as the battery cell accounting for 40%, the circuit board accounting for 30%, and the heat dissipation components accounting for 30%), and then the remaining service life is estimated. For example, if the overall aging is 60%, the remaining service life is 40%.

[0039] The tiered early warning mechanism sets up three response levels based on the remaining maintenance life and the degree of data anomaly: Level 1 warning corresponds to a lifespan > 30% and no key parameter anomalies, the module's green indicator light flashes, a "routine inspection reminder" is pushed locally, and status data is pushed to the backend remotely via 4G / 5G; Level 2 warning corresponds to a lifespan of 10%-30% or a slight anomaly in a single key parameter (such as voltage fluctuations slightly exceeding the threshold), the yellow indicator light stays on and intermittently beeps, a "core component special inspection reminder" is pushed, and the module is marked as a "key focus object" in the backend; Level 3 warning corresponds to a lifespan < 10% or multiple parameter anomalies, the red indicator light flashes rapidly and continuously beeps, an "immediate replacement reminder" is pushed, and remote automatic switching to Beidou communication ensures that early warnings are delivered to remote areas, comprehensively covering the needs from routine maintenance to emergency intervention.

[0040] In some embodiments of the portable power module lifespan prediction and early warning method provided in this application, simultaneous recording of full-cycle data for fault tracing and model optimization includes: The built-in storage module records monitoring data, early warning records, and maintenance operation information for each charge and discharge cycle, and sets a data retention period. Maintenance personnel can read the data from the built-in storage module through a dedicated terminal for fault root cause analysis and maintenance process optimization. The digital model automatically stores prediction process data and regularly updates the core component aging evolution law database to improve prediction accuracy.

[0041] Full-cycle data recording is achieved through the module's built-in storage module. This module records multi-dimensional monitoring data (such as voltage fluctuation curves, ambient temperature and humidity), early warning records (early warning level, trigger time), and maintenance operation information (maintenance items, replaced parts) for each cycle, using "charge and discharge cycles" as the unit. The data retention period is set to 2 years. The design of retaining data for 2 years not only meets the traceability requirements of the average service life of the module (1-2 years) in field operations, but also avoids excessive data occupation of storage resources. Maintenance personnel can read the data through a dedicated terminal (such as a tablet with encryption function). For example, if a module frequently triggers a level 2 early warning, historical temperature data can be reviewed to determine whether long-term high temperature has accelerated aging, providing a basis for fault root cause analysis and maintenance process optimization.

[0042] Model optimization is based on stored prediction process data. The module automatically calculates the deviation between historical predicted lifespan and actual aging results monthly. If the average deviation exceeds 10%, the aging evolution law database is updated. For example, if the actual aging rate of a batch of battery cells is 20% faster than recorded in the database, the aging cycle data and acceleration coefficient of the corresponding battery cells are adjusted. At the same time, the weight calculation logic in the model is optimized. If it is found that the aging of heat dissipation components has a greater impact on the overall lifespan than expected, its functional weight is increased. Through the closed loop of "data recording - deviation analysis - parameter update", the reliability of model prediction is continuously improved.

[0043] In some embodiments of the portable power module lifespan prediction and early warning method provided in this application, the operation simulation scenario generated based on monitoring data and linked in real time with the actual operation scenario includes: Based on scene modeling algorithms, geospatial data of the work area is imported to accurately map the real-time location of the power module and associate it with surrounding environmental interference sources. Based on physical dynamic feature data, the appearance of the power module is visualized in the operation simulation scenario, and potential damage areas are marked. Combined with dynamic data, the movement trajectory and stress of the power module are simulated, and abnormal data triggers priority updates of the operation simulation scenario, giving priority to feedback of abnormal states.

[0044] The core of dynamically updated simulation scenarios is "visual presentation + priority response to anomalies." After importing geographical data of the work area, the scenario modeling algorithm not only accurately maps the real-time location of the module but also correlates it with the distribution of surrounding environmental interference sources. For example, it marks high-temperature areas and strong vibration areas within 500 meters of the module, using different colors to distinguish the interference level, allowing maintenance personnel to intuitively understand the risk environment in which the module is located. Based on physical dynamic feature data, the scenario presents the module's appearance status in the form of a 3D model. For example, cracks in the outer shell are highlighted in red, and oxidized areas of the interfaces are marked in yellow. At the same time, it displays quantitative information such as the damage area and depth, avoiding subjective misjudgments in traditional visual inspections.

[0045] The scenario update adopts a "regular update + anomaly priority" mechanism. Under normal operating conditions, data is synchronized every 5 seconds. When abnormal data is detected (such as a sudden drop in voltage or a sudden rise in temperature), the scenario immediately triggers a priority update, and completes the synchronous display of abnormal parameters and related status within 0.5 seconds. For example, if the voltage suddenly drops to 80% of the rated value, the scenario will simultaneously display the input power fluctuation curve, the current temperature of the circuit board, and mark possible faulty components (such as voltage regulation chips), helping maintenance personnel to quickly locate risk points and improve emergency response efficiency.

[0046] In some embodiments of the portable power module lifespan prediction and early warning method provided in this application, triggering graded early warnings based on remaining maintenance lifespan and the degree of data anomaly includes: Import the geographic information and preset action route data of the power module for subsequent operations; reproduce the subsequent geographic environment and action route in the running simulation scenario, and predict the aging rate change of core components in combination with the database; calculate the remaining maintenance life of multiple subsequent time nodes based on the predicted rate, generate the life change trend curve, and trigger the forward warning in advance if the predicted life will enter the warning range.

[0047] Based on geographic location and operational routes, this lifespan prediction method overcomes the limitations of existing technologies that "can only predict in real time, but cannot predict in advance." Its core is to combine parameters of subsequent operational scenarios to achieve "future lifespan extrapolation." First, it imports geographic information data of the operational area, including terrain (mountains, plains), altitude (high altitude reduces heat dissipation efficiency), climate (high temperature and humidity, low temperature and dryness), and the location of environmental interference sources (such as strong electromagnetic fields near factories). At the same time, it acquires preset operational route data, covering the movement path (such as the specific route from camp A to work point B) and stop nodes (the duration of operation at each node, such as a 4-hour stop at point C), ensuring the completeness of scenario reproduction.

[0048] In the simulated operation scenario, the future operating state of the module is reproduced according to the subsequent geographical environment and action route. Combined with the aging evolution law database, the aging rate at different stages is predicted. For example, if the module will enter a high temperature and high humidity area in the next 24 hours, the aging rate of the battery cells is predicted to increase by 40%, and the aging rate of the circuit board will increase by 35%. If it enters a strong vibration mountain area after 72 hours, the aging rate of the solder joints is predicted to increase by 50%. Based on these predicted rates, the remaining maintenance life after 24 hours, 72 hours, and one week is calculated, and the life change trend curve is generated. If it is predicted that the life will drop from the current 35% to 28% after 72 hours (entering the second-level warning range), a forward-looking warning is triggered 24 hours in advance, and a prompt to "prepare core component testing tools in advance" is pushed, providing a time window for advance resource scheduling for field operations.

[0049] In some embodiments of the portable power module lifespan prediction and early warning method provided in this application, triggering graded early warnings based on remaining maintenance lifespan and the degree of data anomaly includes: When multiple power modules are in the same working area, the digital models of each module share data through remote communication; based on the shared data, an aging correlation model of the regional power cluster is constructed to analyze the mutual influence of aging between modules; when any module in the cluster triggers an early warning, the correlation model is used to predict the impact on other modules and push collaborative maintenance suggestions.

[0050] Cross-module collaborative early warning addresses the issue of "single-module early warning not considering cascading risks" in multi-module cluster operation scenarios (such as 10 modules being powered simultaneously in military field support). Its core is "data sharing + correlation modeling". When multiple modules are in the same operating area (such as within a 1-kilometer radius), the digital models of each module achieve data interoperability through remote communication (4G / 5G or BeiDou). They share regional environmental data (such as overall temperature and humidity, average vibration intensity), motion characteristic data (such as whether the movement trajectories of each module intersect), and aging status data (such as the current aging percentage of each module), avoiding risk misjudgment caused by isolated data.

[0051] Based on shared data, an aging correlation model for the regional power cluster is constructed. This model analyzes the mutual influence of aging between modules. For example, if a module's internal short circuit causes a 5°C increase in the surrounding ambient temperature, the model will calculate the accelerated aging effect of this high temperature on the cells of other modules within 50 meters (e.g., an increase in aging rate of 15%). If a module bears an excessively high load (exceeding 120% of its rated load), the model will predict the load sharing of other modules after its failure and the resulting accelerated aging (e.g., a 30% increase in load leading to a 25% increase in aging rate). When any module in the cluster triggers an early warning, such as module A triggering a level 3 warning, the model will predict that its failure may cause a sudden increase in the load on modules B and C, and will proactively push out a collaborative maintenance suggestion to "transfer the load of modules B and C to other modules" to avoid cascading failures that could lead to regional power outages.

[0052] In some embodiments of the portable power module lifespan prediction and early warning method provided in this application, simultaneous recording of full-cycle data for fault tracing and model optimization includes: Based on the time-series characteristics of monitoring data and changes in scenario parameters, an aging causative factor feature library is constructed. By comparing real-time data with the causative factor feature library through machine learning algorithms, the core aging causative factors and their contribution ratios are identified.

[0053] While outputting the remaining maintenance life, it generates a cause diagnosis report, clearly identifies prevention recommendations to delay aging and optimize subsequent operations.

[0054] The core of the intelligent diagnostic function for aging causes is "feature matching + cause quantification". The digital model first constructs an aging cause feature library based on the time-series change characteristics of multi-dimensional monitoring data, combined with environmental and dynamic parameter changes in the simulation scenario. This feature library covers typical data characteristics of four types of causes: environmental stress (high temperature, high humidity, strong vibration), mechanical damage (shell cracking, cell displacement), electrical load (overcharging and over-discharging, overload), and storage loss (long-term idleness, low-temperature storage). For example, the characteristic of high temperature cause is "continuous 4 hours of temperature > 40℃, cell aging rate increases by more than 30%", and the characteristic of mechanical damage cause is "after vibration acceleration > 18m / s², circuit board aging rate increases sharply".

[0055] During diagnosis, the model uses machine learning algorithms (such as random forest) to compare real-time multi-dimensional monitoring data with a feature library of causes, calculating similarity. If the similarity of a certain type of cause exceeds 85%, it is identified as a core cause, and its contribution percentage is quantified. For example, in the case of accelerated aging of a module, high temperature contributes 60%, overcharging contributes 30%, and other factors contribute 10%. While outputting the remaining service life, a cause diagnosis report is generated, clearly stating that "high temperature is the main cause, and it is recommended to strengthen heat dissipation protection." At the same time, it provides data support for the selection of subsequent operating scenarios, such as avoiding deployment of the module in high-temperature areas, thus delaying aging from the source and improving the module's service life.

[0056] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A portable power module life prediction and early warning method, characterized in that, Comprise: Collecting multi-dimensional monitoring data of the power module by integrated sensor group, which at least covers the operating state, field environment and physical dynamic characteristic data of the power module; Generating an operating simulation scenario in real-time linkage with the actual working scenario based on the monitoring data, which dynamically maps the real-time working condition and spatial position of the power module; Building a digital model matching the actual power module specifications in the operating simulation scenario, which has a built-in core component aging evolution law database; Real-time inputting the monitoring data into the digital model, comprehensively judging the aging process and calculating the remaining maintenance life of the power module by combining the environmental and dynamic characteristics of the core component aging acceleration, triggering a graded warning according to the remaining maintenance life and the degree of data anomaly, pushing maintenance prompts through local alarm and remote dual-mode communication, and recording full-cycle data for fault tracing and model optimization.

2. The portable power module life prediction and warning method of claim 1, wherein, The multi-dimensional monitoring data of the power module collected by the integrated sensor group comprises: The operating state covers the charge-discharge cycle number, output voltage stability, internal core component temperature, charge-discharge rate, external input power fluctuation and idle storage time of the power module, which is collected by an adaptive sensor unit; The field environment covers the temperature and humidity, vibration intensity of the working area, which is collected by sensors deployed at specific positions of the module; the characteristic data covers the appearance integrity, surface damage and dynamic parameters during movement of the key parts of the power module, which is collected by corresponding sensor units.

3. The portable power module life prediction and warning method of claim 1, wherein, The operating simulation scenario in real-time linkage with the actual working scenario based on the monitoring data comprises: Pretreating the monitoring data, eliminating invalid data, eliminating transient interference and standardizing different dimension data; setting a regular self-calibration mechanism for sensors, correcting the collection parameters by built-in standard reference source at a preset period to avoid data drift; Using scenario modeling algorithm to fuse pretreated data and geographic spatial information, building a dynamically updated operating simulation scenario to ensure consistency with the actual scenario.

4. The portable power module life prediction and warning method of claim 1, wherein, The digital model has a built-in core component aging evolution law database, which comprises: The database is built based on a large number of actual application cases, containing aging period data of battery, circuit board, heat dissipation components and other components under different conditions, and presetting multi-dimensional aging acceleration coefficients; the digital model and the operating simulation scenario establish a real-time data interaction link to ensure that the digital model parameters dynamically match the actual working condition of the power module; the digital model has data learning ability, which can iteratively optimize the related data and acceleration coefficients in the core component aging evolution law database based on historical prediction bias.

5. The portable power module life prediction and warning method of claim 1, wherein, The graded warning triggered according to the remaining maintenance life and the degree of data anomaly comprises: The digital model matches real-time data with database historical data, calls corresponding acceleration coefficient calculation core component aging percentage, weights overall aging degree and calculates remaining maintenance life; the hierarchical early warning mechanism at least includes three levels of early warning, different early warning levels correspond to different remaining maintenance life intervals and data anomaly degrees, and when triggered, maintenance guidance is pushed, and the remote communication supports multi-mode automatic switching to adapt to field conditions.

6. The portable power module life prediction and warning method of claim 1, wherein, The simultaneous recording of full-cycle data is used for fault tracing and model optimization, including: The built-in storage module records the monitoring data, early warning records and maintenance operation information of each charge and discharge cycle, and sets a data retention period; maintenance personnel can read the data in the built-in storage module through a special terminal for fault root analysis and maintenance process optimization; the digital model automatically stores prediction process data and regularly updates the core component aging evolution law database to improve prediction accuracy.

7. The portable power module life prediction and warning method of claim 3, wherein, The running simulation scene generated based on the monitoring data fitting includes real-time linkage with actual operation scenarios, including: Based on the scene modeling algorithm, the geographic spatial data of the operation area is imported, the real-time position of the power module is accurately mapped, and the surrounding environmental interference sources are associated; based on the physical dynamic characteristic data, the appearance state of the power module is visually presented in the running simulation scene, and the potential damage area is marked; the motion trajectory and stress condition of the power module are simulated in combination with dynamic data, and abnormal data triggers priority update of the running simulation scene, and abnormal state is preferentially fed back.

8. The portable power module life prediction and warning method of claim 1, wherein, The hierarchical early warning triggered according to the remaining maintenance life and the data anomaly degree includes: The geographic information and preset action route data of the power module in subsequent operation are imported; the subsequent geographic environment and action route are reproduced in the running simulation scene, and the database predicted core component aging rate change is combined; the remaining maintenance life at multiple time nodes in the future is calculated based on the predicted rate, a life change trend curve is generated, and if the predicted life will enter the early warning interval, a forward-looking early warning is triggered in advance.

9. The portable power module life prediction and warning method of claim 1, wherein, The hierarchical early warning triggered according to the remaining maintenance life and the data anomaly degree includes: When multiple power modules are in the same operation area, the digital models of each module share data through remote communication; based on the shared data, a regional power cluster aging correlation model is constructed, and the mutual influence law of the modules is analyzed; when any of the modules in the cluster triggers an early warning, the influence on other modules is predicted in combination with the correlation model and a collaborative maintenance suggestion is pushed.

10. The portable power module life prediction and warning method of claim 1, wherein, The simultaneous recording of full-cycle data is used for fault tracing and model optimization, including: Based on the timing characteristics and scene parameter changes of the monitoring data, a feature library of aging inducements is constructed; through machine learning algorithm, real-time data is compared with the inducement feature library to identify core aging inducements and contribution proportions; While outputting the remaining maintenance life, an inducement diagnosis report is generated to clearly specify inhibition suggestions to delay aging and optimize subsequent operation.

Images