Self-adaptive evaluation method and system for service life of battery of wireless vibration sensor

Abstract

The invention discloses a wireless vibration sensor battery life self-adaptive evaluation method and system, and belongs to the technical field of wireless vibration sensor battery predictive maintenance. In order to solve the technical problem that the battery residual life of a sensor in different working modes and wide-temperature-zone environments cannot be accurately evaluated due to dependence on voltage monitoring in the prior art, the invention provides a solution integrating accurate metering, mechanism modeling and intelligent learning. According to the method, the actual power consumption of the battery is directly and accurately metered through the coulombmeter module, and a traditional voltage monitoring mode is replaced; a dynamic power consumption model for deeply associating sensor configuration information with environment temperature is constructed, and dynamic correction of power consumption evaluation is realized; and a three-dimensional self-learning mechanism of transverse cluster comparison, longitudinal time sequence iteration and working condition closed-loop verification is designed, so that the system can be self-adaptive to battery aging and hardware drift for a long time. Finally, in combination with future temperature prediction, the high-precision remaining working time of the battery is output.

Description

Technical Field

[0001] This invention relates to the field of predictive maintenance technology for wireless vibration sensor batteries, and more particularly to an adaptive assessment method and system for wireless vibration sensor battery life. Background Technology

[0002] Wireless vibration sensors, as key sensing nodes in the Industrial Internet of Things (IIoT), are widely used in monitoring the operational status of rotating machinery such as motors and bearings. Due to the complex environments of industrial sites and the often lack of wired power supply, these sensors are typically powered by disposable batteries and must operate stably for extended periods in a wide temperature range of -40℃ to 85℃. Therefore, accurately assessing the remaining battery life is crucial for developing reasonable equipment maintenance plans and preventing monitoring interruptions due to battery depletion.

[0003] However, existing battery remaining life assessment technologies still have significant shortcomings. For example, Chinese patent announcement CN114216558B discloses a method and system for predicting the remaining life of a wireless vibration sensor battery. This scheme mainly collects battery voltage, temperature, and timestamps, and uses machine learning regression algorithms to build a prediction model. However, this patented technology has the following limitations: First, it mainly relies on voltage statistical characteristics to characterize the battery state, while industrial lithium batteries (such as lithium-thionyl chloride batteries) have a flat discharge plateau characteristic, and the relationship between voltage and capacity is non-linear. Relying solely on voltage is insufficient to accurately reflect the remaining charge in the later stages of discharge. Second, this scheme adopts a "data-driven" black-box approach, ignoring the decisive influence of the sensor's own "configuration information" (such as acquisition interval, upload interval, and data length) on power consumption, i.e., it does not establish a physical correlation between "operating mode and power consumption." Finally, its model lacks an online adaptive calibration mechanism for specific operating conditions. When the sensor adjusts its configuration parameters or long-term operation causes hardware characteristic drift, the model cannot automatically correct itself, resulting in a significant increase in remaining life assessment errors, making it difficult to meet the needs of high-precision predictive maintenance in industrial settings. Summary of the Invention

[0004] The purpose of this invention is to solve the technical problem of large errors in the remaining life assessment of existing wireless vibration sensors due to the mismatch between the fixed power consumption model and actual operating conditions under different working modes and wide temperature ranges. By introducing a coulomb counter module to accurately measure the actual power consumption and constructing a dynamic power consumption model that deeply correlates sensor configuration information (including acquisition interval and upload interval) with ambient temperature, the power consumption characteristics of the sensor under different operating modes can be accurately characterized and dynamically corrected.

[0005] The purpose of this invention is to overcome the technical shortcomings of existing technologies that rely solely on data regression and lack physical mechanism support, thus failing to effectively address battery aging and hardware drift. By establishing a capacity correction logic that includes segmented temperature decay and dual aging factors, and combining it with a three-dimensional self-learning mechanism of horizontal cluster comparison and vertical time-series iteration, the evaluation model can be online adaptively calibrated, ensuring that the system maintains high-precision evaluation capabilities during long-term operation and hardware aging.

[0006] The purpose of this invention is to address the problem that existing technologies rely solely on static assessments based on historical data, lacking forward-looking considerations for future operating conditions. By introducing time-series analysis algorithms to predict future ambient temperature trends and adjusting the dynamic power consumption model accordingly to estimate future expected power consumption, this invention provides users with predictions of remaining battery operating time based on actual future operating conditions, effectively avoiding unplanned downtime caused by sudden environmental changes or sudden drops in battery power.

[0007] This invention proposes an adaptive assessment method for the battery life of a wireless vibration sensor. The method includes: acquiring the sensor's coulomb counter charge, ambient temperature, and configuration information including acquisition and upload intervals; determining a corresponding temperature decay factor based on the ambient temperature and correcting the battery's rated capacity according to its aging level to obtain the current actual usable capacity; matching the baseline theoretical power consumption based on the configuration information and dynamically correcting the baseline theoretical power consumption using the actual power consumption calculated from the coulomb counter charge, thus constructing a dynamic power consumption model reflecting changes in operating conditions; predicting future ambient temperatures based on historical temperature data and adjusting the dynamic power consumption model accordingly to obtain the future predicted power consumption; and calculating the remaining battery operating time by combining the current actual usable capacity with the current remaining power. This method integrates coulomb counter measurement, configuration, and temperature information to establish an assessment closed loop of "precise measurement - dynamic modeling - forward-looking prediction," fundamentally solving the technical problem of inaccurate power assessment under varying operating conditions in traditional voltage monitoring methods.

[0008] Preferably, this method divides the operating temperature range of -40℃ to 85℃ into several preset temperature intervals, determines the target interval to which the current ambient temperature belongs, and calculates the temperature decay factor using the linear correction coefficient corresponding to the target interval. By segmenting the wide temperature range, this method effectively quantifies the nonlinear decay of battery capacity under extreme high and low temperatures, significantly improving the accuracy of capacity assessment across the entire temperature range (especially from -40℃ to 85℃).

[0009] Preferably, this method obtains the battery's cumulative charge-discharge cycle count and usage time; calculates the active material loss caused by the cumulative charge-discharge cycle count using the cycle aging factor, and calculates the self-discharge loss caused by the usage time using the calendar aging factor, combining the two to obtain the aging degradation factor; and multiplies the rated capacity, temperature degradation factor, and aging degradation factor to obtain the current actual usable capacity. This dual correction mechanism of cycle aging and calendar aging enables the model to dynamically reflect the battery's capacity degradation throughout its entire lifespan, overcoming the technical deficiency of existing technologies that ignore the impact of long-term static self-discharge.

[0010] Preferably, this method queries a preset configuration-power consumption mapping table based on the acquisition interval and the upload interval to obtain the theoretical reference power consumption of the current configuration at standard temperature, and then uses the current ambient temperature to perform temperature compensation on the theoretical reference power consumption to obtain the initial theoretical average power consumption. By establishing the configuration-power consumption mapping table and combining it with temperature compensation, an accurate theoretical power consumption benchmark is provided for sensors in different operating modes, achieving a deep integration of power consumption assessment and the actual operating logic of the sensor.

[0011] Preferably, this method determines whether the number of sensor operation cycles has reached a preset stability threshold. If the stability threshold has not been reached, the initial theoretical average power consumption is directly determined as the dynamic power consumption of the current cycle. If the stability threshold has been reached, the actual average power consumption of multiple historical cycles is calculated, and the initial theoretical average power consumption and the actual average power consumption are weighted and fused according to preset weights. The fused result is then determined as the dynamic power consumption of the current cycle. This two-stage dynamic power consumption determination strategy of "theoretical preset + measured optimization" enables the system to operate stably in the initial stage and continuously approach real-world operating conditions as data accumulates, improving the model's adaptability and reliability.

[0012] Preferably, this method divides the current sensor into a sensor cluster with the same configuration information and within the same temperature range; calculates the average power consumption of all sensors in the sensor cluster as the cluster benchmark; if the dynamic power consumption of the current sensor exceeds the preset deviation range of the cluster benchmark, the sensor power consumption is determined to be abnormal, and the theoretical benchmark power consumption of the current sensor is corrected using the cluster benchmark. Through horizontal cluster comparative learning, the abnormal power consumption of individuals is identified and corrected using group data characteristics, effectively eliminating the evaluation bias caused by hardware discreteness and enhancing the robustness of the system.

[0013] Preferably, after detecting a battery replacement event, this method obtains the actual operating time before the battery replacement; calculates the evaluation error between the actual operating time and the estimated remaining operating time before the system replacement; if the evaluation error exceeds a preset tolerance threshold, the calculation parameters of the temperature decay factor or the theoretical reference power consumption of the configuration information are adjusted according to the positive or negative direction of the error. This closed-loop verification and parameter calibration mechanism based on battery replacement events enables the system to learn from actual results, continuously optimize model parameters, and ensure the stability of long-term evaluation accuracy.

[0014] This invention proposes an adaptive battery life assessment system for wireless vibration sensors. This system is applied to the aforementioned adaptive battery life assessment method for wireless vibration sensors. The system includes: a coulomb counter module whose power input is electrically connected to the battery; a coulomb counter module whose power output is electrically connected to a power management unit; and a coulomb counter module whose data communication terminal is connected to a microcontroller. A battery temperature acquisition unit's data output is also connected to the microcontroller. The power management unit's voltage output is electrically connected to the microcontroller, the wireless communication unit, and the signal acquisition unit, respectively. The microcontroller is connected to the coulomb counter module and the battery temperature acquisition unit via a data bus. The microcontroller stores a sensor configuration information table containing acquisition intervals and upload intervals. This hardware system architecture, through the series design of the coulomb counter modules, achieves direct, end-to-end accurate measurement of the total battery power consumption, providing an indispensable hardware foundation for high-precision assessment.

[0015] Preferably, the system also includes a signal acquisition unit and a data buffer unit. The power supply terminal of the signal acquisition unit is electrically connected to the voltage output terminal of the power management unit, and the signal output terminal of the signal acquisition unit is physically connected to the input interface of the microcontroller. The signal acquisition unit includes an accelerometer for sensing vibration signals. The data buffer unit is electrically connected to both the power management unit and the microcontroller's storage expansion interface, and is a non-volatile memory chip. Integrating the signal acquisition and data buffer units ensures that critical configuration data and historical records are not lost during power outages while supporting the core vibration sensing function, thus improving the system's practicality and reliability.

[0016] Preferably, the coulomb counter module is connected in series between the positive terminal of the battery and the input terminal of the power management unit. The coulomb counter module integrates a current sampling resistor and an analog-to-digital converter (ADC). The current sampling resistor is connected in series in the power supply circuit, and the digital signal output pin of the ADC is connected to the communication interface of the microcontroller via an I2C or SPI bus. This series integration design of the coulomb counter module and the power supply circuit, along with the digital interface, enables non-destructive, real-time monitoring of a wide current range from microamperes to milliamperes, providing a high-precision raw data source for dynamic power consumption models.

[0017] The present invention has the following beneficial effects: 1. This invention solves the problem of power consumption model distortion under multiple operating conditions, achieving deep decoupling and dynamic adaptation between the evaluation model and sensor business logic. By introducing a coulomb counter module to accurately measure actual power consumption, this invention constructs a dynamic power consumption model that deeply correlates sensor configuration information (including acquisition interval and upload interval) with ambient temperature. Unlike existing technologies that rely solely on voltage statistical characteristics or fixed power consumption parameters, this invention can quickly match new theoretical benchmarks using configuration information and dynamically correct them by combining measured data when the power consumption characteristics change abruptly due to sensor adjustments to operating modes (such as increasing the acquisition frequency). This mechanism effectively overcomes the limitation of single voltage monitoring methods in sensing changes in load conditions, significantly improving the system's evaluation robustness under complex and variable industrial site configurations.

[0018] 2. This invention overcomes the limitations of purely data-driven models lacking physical mechanism support, ensuring high-precision capacity characterization in extreme temperature ranges and throughout the entire battery lifespan. It establishes a physical capacity correction logic that integrates segmented temperature decay and dual aging factors. By dividing the operating temperature range into different segments and applying linear correction coefficients to each, the nonlinear capacity decay characteristics of the battery under extreme high and low temperatures are accurately fitted. Simultaneously, by combining cycle aging and calendar aging calculations to determine the comprehensive aging factor, it overcomes the shortcomings of traditional methods that only consider the number of cycles while ignoring long-term self-discharge during static storage. This allows the system to maintain a highly reliable usable capacity assessment capability even in the mid-to-late stages of battery discharge and during periods of drastic environmental temperature fluctuations.

[0019] 3. A three-dimensional self-learning mechanism with self-evolutionary capabilities has been established, eliminating the cumulative impact of individual hardware differences and long-term drift on evaluation accuracy. This invention achieves online adaptive calibration of model parameters through horizontal cluster comparison and vertical time-series iteration. The system can utilize the characteristics of group data under the same configuration and environment to identify and correct individual anomalies, while automatically adjusting the aging coefficient through long-term trend analysis. This closed-loop feedback mechanism solves the problem that existing technologies cannot adapt to the drift of hardware characteristics over time once the model parameters are set, significantly reducing the system's dependence on manual on-site calibration and ensuring the stability of the evaluation system during long-term operation.

[0020] 4. This invention represents a leap from "current status assessment" to "proactive prediction," effectively mitigating the risk of unplanned downtime caused by sudden environmental changes. It introduces a future environmental temperature prediction mechanism based on time-series analysis, and uses this to calculate expected future power consumption. By extending the assessment perspective from current remaining battery capacity to energy consumption under future operating conditions, the system can identify in advance the risk of reduced battery performance or surged power consumption due to sudden drops in future temperatures. This provides users with dynamic remaining battery life predictions based on the actual future operating environment, significantly improving the scientific rigor and safety of maintenance planning compared to existing technologies that rely solely on linear extrapolation from historical average data. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0022] Figure 2 This is a system module architecture diagram of the present invention. Detailed Implementation

[0023] Example 1 according to Figure 1 As shown, this invention provides a method for assessing the remaining battery operating time based on a wireless vibration sensor. This method is mainly applied to the monitoring of rotating machinery in industrial IoT scenarios. Wireless vibration sensors are typically deployed in various complex industrial environments, and their power supply batteries face challenges such as wide temperature range variations, long-cycle operation, and intermittent high-concurrency data transmission. To achieve high-precision lifespan prediction, this embodiment first performs a data acquisition step. The system works collaboratively through a coulomb counter module, a temperature sensor, and a microcontroller integrated into the sensor hardware circuit. The coulomb counter module is connected in series between the battery and the load circuit, performing real-time integration calculations on the current flowing through it to obtain the battery's cumulative charge data. Simultaneously, the temperature sensor collects the current ambient temperature data according to a preset sampling frequency. More importantly, the system synchronously reads the sensor's current configuration information, which is a core parameter determining the sensor's power consumption baseline. This configuration information specifically includes the data acquisition interval (i.e., how often to wake up for vibration acquisition), the data upload interval (i.e., how often to activate the wireless radio frequency module to send data packets), and the length of data acquired per acquisition. To eliminate the interference of environmental noise on the data, the system smooths the raw temperature sequence data after acquisition. For example, it uses a filtering algorithm based on Gaussian distribution weights to remove instantaneous temperature jumps, thereby obtaining a smooth average temperature that better reflects the current environmental thermal state, providing a stable and reliable input basis for subsequent model calculations.

[0024] After completing the acquisition and preprocessing of basic data, the system proceeds to the actual power consumption calculation step. This step aims to obtain the actual energy consumption of the sensor in the real physical world, serving as the benchmark for subsequent calibration of the theoretical model. The system calculates the difference between the cumulative charge of the coulomb counter at the end of the current cycle and the end of the previous cycle, thereby accurately determining the actual power consumption within the current cycle. This coulomb counter integration-based measurement method is fundamentally different from the existing method of simply relying on the battery voltage curve to infer the power consumption. Because industrial-grade lithium batteries (such as lithium thionyl chloride batteries) have an extremely flat discharge voltage plateau, the voltage remains almost constant for most of their lifespan, only dropping sharply when the charge is about to be depleted. Therefore, the voltage method is extremely insensitive in the middle of the discharge process. However, the charge difference calculation method used in this embodiment can sensitively capture the minute amount of power consumed by each vibration acquisition and wireless transmission. Regardless of the battery's discharge stage, it can provide objective and linear energy consumption data, providing solid data support for the subsequent correction of the dynamic power consumption model.

[0025] Simultaneously, the system executes the capacity assessment step in parallel. The core of this step lies in constructing a capacity correction model that dynamically reflects the battery's physical characteristics, addressing the problem of existing technologies neglecting the influence of environment and time on the battery's "tank size." The system first determines the temperature degradation factor based on the currently collected ambient temperature. Considering the nonlinear response of battery chemical activity to temperature, this embodiment pre-constructs a piecewise function model covering the entire operating temperature range (e.g., from extremely cold temperatures of tens of degrees below zero to high temperatures of over eighty degrees Celsius). The system determines which preset temperature range the current temperature falls into and calls the corresponding linear correction coefficient to calculate the temperature's impact on capacity. For example, in the low-temperature range, increased electrolyte viscosity leads to decreased ion mobility, and the correction coefficient significantly reduces the effective capacity; while in the high-temperature range, the impact of increased self-discharge is considered. In addition to temperature-based correction, the system also introduces a dual aging correction mechanism, combining the battery's cumulative charge-discharge cycle count and usage time to determine the aging degradation factor. This factor comprehensively quantifies the irreversible loss of active material due to cyclic use (cycle aging) and the loss due to internal micro-short circuits or chemical decomposition caused by long-term static storage (calendar aging). Finally, the system multiplies the battery's factory rated capacity with the aforementioned temperature degradation factor and aging degradation factor to obtain the battery's actual usable capacity under the current operating conditions.

[0026] Based on the actual power consumption and available capacity, the system executes the core dynamic power consumption modeling step. This step aims to address the technical pain point that fixed power consumption models cannot adapt to flexible adjustments in sensor configuration. The system first queries a preset mapping relationship based on the read configuration information (acquisition interval, upload interval, etc.) to match the theoretical baseline power consumption of that configuration under standard conditions. Subsequently, the system compensates for this theoretical baseline power consumption using the current ambient temperature, forming an initial theoretical average power consumption. However, theoretical values ​​often fail to fully cover individual hardware differences and characteristic drift after long-term operation. Therefore, the system introduces dynamic correction logic: after the sensor has reached a certain stable operating period, the system compares the actual power consumption measured by the coulomb counter with the theoretical value. If a deviation is found, the system will update the baseline theoretical power consumption using a weighted fusion algorithm according to a preset weighting strategy. For example, in the early stages of model building, more reliance is placed on theoretical values ​​to maintain stability, while after sufficient data accumulation, more weight is given to the measured values ​​to improve agility. This mechanism enables the power consumption model to "evolve," meaning that when the sensor increases its sampling frequency due to field requirements, or when leakage current increases due to component aging, the model can automatically sense and adjust its parameters to ensure that the power consumption assessment model always matches the actual operating state of the sensor.

[0027] Finally, the system performs a lifespan prediction step, achieving a leap from "current status assessment" to "prospective prediction." To avoid unplanned downtime due to sudden environmental changes in the future (such as a cold wave), this embodiment abandons the traditional approach of simply extrapolating the remaining lifespan using historical average power consumption. The system integrates a time series analysis algorithm to predict the environmental temperature trend within a preset time period based on historical temperature change trends. Then, the system substitutes the predicted future temperature into a dynamically corrected power consumption model to calculate the expected average power consumption of the sensor under future conditions. This calculation method fully considers the increased power consumption that low temperatures may cause (such as increased battery internal resistance leading to longer transmission time) and the capacity loss that high temperatures may cause. Finally, the system combines the current available capacity, the accumulated charge consumed (i.e., the current remaining power), and the predicted future average power consumption to calculate the remaining battery operating time. This result is not only a single time value but can also include a range based on model confidence, providing highly forward-looking and reliable reference for maintenance personnel in industrial fields to formulate replacement plans.

[0028] Example 2 according to Figure 2As shown, this invention provides a battery remaining operating time assessment system based on a wireless vibration sensor. This system overcomes the shortcomings of existing technologies that rely solely on voltage monitoring, leading to inaccurate energy representation, and provides necessary hardware support for the operation of dynamic power consumption models. The system mainly includes a power supply and metering section, an environmental sensing section, a core control section, and a data interaction section. The core control section, composed of a microcontroller (MCU), serves as the system's control center, coordinating the working timing of each module and running complex battery assessment algorithms. The core components of the power supply and metering section include a battery, a coulomb counter module, and a power management unit. The battery, as the sole energy source, provides power to the entire system. To achieve accurate energy measurement, this embodiment introduces a coulomb counter module between the battery and the subsequent load circuit. The power input terminal of the coulomb counter module is directly electrically connected to the positive terminal of the battery via a wire, while its power output terminal is connected to the input terminal of the power management unit. This series configuration ensures that all current output from the battery must flow through the sampling resistor inside the coulomb counter module. This guarantees that both the microampere-level leakage current when the sensor is in sleep mode and the milliampere-level peak current during wireless transmission can be captured and integrated by the coulomb counter module. The coulomb counter module converts the acquired analog current signal into digitized charge data through its internal high-precision analog-to-digital converter. This data is then connected to the microcontroller's communication interface via a data communication terminal (such as an I2C or SPI interface), transmitting the accumulated charge consumption data back to the microcontroller in real time. This provides a realistic physical energy consumption basis for subsequent remaining time assessment.

[0029] The power management unit (PMU) plays a crucial role in voltage regulation and power distribution within the system. Its input receives the voltage from the coulomb counter module, which is then processed by internal voltage regulation circuitry (such as a low-dropout linear regulator (LDO) or a DC-DC converter) to output a stable operating voltage. The PMU's output terminals are electrically connected to the microcontroller, wireless communication unit, and signal acquisition unit, forming a complete power supply network. This design ensures that even as battery voltage fluctuates with capacity decay, each functional module can still operate stably within its rated voltage range, avoiding measurement errors or communication failures caused by voltage fluctuations. Particularly for high-power loads such as the wireless communication unit and signal acquisition unit, the PMU provides current support with excellent transient response, guaranteeing system stability during data transmission and vibration acquisition. Furthermore, the PMU is controlled by the microcontroller and can independently manage the power supply to certain peripheral modules according to different system operating modes (such as acquisition mode and sleep mode), thereby minimizing the system's static power consumption.

[0030] The environmental sensing section mainly consists of a battery temperature acquisition unit and a signal acquisition unit. The battery temperature acquisition unit is typically attached to the battery surface or integrated near the battery pack, used to monitor the battery's operating ambient temperature in real time. Its data output is directly connected to the microcontroller, transmitting the acquired analog or digital temperature signal to the microcontroller. This hardware design is the physical basis for implementing the "temperature-capacity" correction model, enabling the microcontroller to acquire real-time ambient temperature data and then call a pre-stored temperature decay piecewise function to correct the battery's actual usable capacity, solving the problem of inaccurate capacity assessment under extreme temperatures. The signal acquisition unit, as the core sensing component of the vibration sensor, includes a high-sensitivity accelerometer. Its power supply is provided by the power management unit, and its signal output is physically connected to the microcontroller's input interface. This unit is responsible for capturing vibration signals from mechanical equipment and converting the physical vibration into electrical signals, which are then transmitted to the microcontroller for processing. It is worth noting that the operating frequency and sampling duration (i.e., configuration information) of this signal acquisition unit directly determine the system's power consumption level. Its connection to the microcontroller is not only a data path but also the physical source of the "configuration information" in the power consumption model.

[0031] The wireless communication unit and data caching unit constitute the system's data interaction and storage components. The wireless communication unit (such as a LoRa or NB-IoT module) connects to the microcontroller via a serial port or SPI bus, responsible for sending processed power data, temperature data, and remaining time assessment results to a remote server or cloud platform. Its transmission frequency (upload interval) is another key configuration parameter affecting power consumption. To accommodate potential communication interruptions or data backlogs in industrial environments, the system is also equipped with a data caching unit, which is electrically connected to both the power management unit and the microcontroller's storage expansion interface. The data caching unit typically uses a non-volatile memory chip (such as Flash or EEPROM) to temporarily store collected vibration data and system operation logs. More importantly, while the microcontroller interacts with each module via the data bus, its internally integrated non-volatile memory contains a sensor configuration information table including the acquisition interval and upload interval, as well as a temperature decay segmentation function table and a dynamic power consumption correction algorithm. The combination of these hardware and software features enables the microcontroller to dynamically adjust the power consumption model parameters based on the stored configuration information and the real-time coulomb counter data, thereby achieving perfect support of the hardware architecture for the "mechanism + data" dual-drive evaluation method.

[0032] Example 3 This embodiment focuses on the specific implementation logic of the core algorithm in the battery remaining working time assessment process. When determining the actual usable capacity of the battery, the system does not simply call the factory nominal value, but adopts a comprehensive calculation strategy that integrates temperature nonlinear decay and multi-dimensional aging effects. Specifically, the processor first reads the battery's rated capacity value, then calculates the temperature decay factor and aging decay factor, and multiplies the rated capacity with these two factors to obtain the actual usable capacity under the current operating conditions. The temperature decay factor is determined based on a pre-constructed piecewise linear function covering the entire operating temperature range. The system divides the operating temperature into several continuous intervals, such as extremely low temperature intervals, low temperature intervals, normal temperature intervals, and high temperature intervals. For each temperature interval, different slope coefficients and intercept constants are preset. When the collected smoothed average temperature falls into a specific interval, the system uses the corresponding linear relationship of that interval to calculate the dimensionless temperature decay coefficient. For example, in extremely low temperature environments, this coefficient reflects the physical characteristic that capacity decreases sharply with decreasing temperature; while in high temperature environments, this coefficient reflects a specific decay trend of capacity with increasing temperature. This segmented modeling approach can accurately fit the actual chemical activity of a battery under non-ideal temperatures.

[0033] Regarding the calculation of the aging degradation factor, this embodiment employs a composite model integrating cyclic aging and calendar aging. The system acquires the battery's cumulative charge-discharge cycle count and battery usage time. For the cyclic aging portion, the system uses the product of a preset cyclic aging coefficient and the cumulative cycle count to characterize the capacity reduction caused by the loss of active materials. For the calendar aging portion, considering that the self-discharge characteristics of a primary battery typically change non-linearly over time, the system uses a calculation logic based on a logarithmic function. That is, it uses natural logarithm calculation to process the ratio of the used time to a specific time base, and combines it with the calendar aging coefficient to quantify the self-discharge loss. Finally, the system subtracts the aforementioned cyclic aging loss and calendar aging loss from the unit base value to obtain the aging degradation factor reflecting the current battery health. When the daily average power consumption change rate is detected to exceed a preset abnormal threshold, the system will also automatically increase the aforementioned cyclic aging coefficient and calendar aging coefficient to adapt to the actual situation of accelerated battery aging or hardware performance degradation.

[0034] In constructing the dynamic power consumption model, this embodiment adopts a two-stage strategy combining "initial theoretical preset" and "later field optimization." In the initial stage of sensor operation, due to a lack of sufficient historical measured data, the system primarily calculates theoretical power consumption based on configuration information. The system internally stores theoretical baseline power consumption values ​​for different configuration categories (e.g., low-power, medium-power, and high-power) at standard temperatures. The system matches the corresponding baseline value based on the current acquisition and upload intervals and adjusts it using a correction coefficient based on the current temperature to obtain the initial theoretical average power consumption. This temperature correction coefficient reflects the impact of temperature on the power consumption characteristics of electronic components; for example, the power consumption coefficient increases significantly at low temperatures. When the number of sensor operation cycles reaches a preset stable threshold (e.g., five cycles), the system enters the dynamic optimization stage. At this point, the system no longer relies solely on theoretical values ​​but calculates the average power consumption measured by the coulomb counter over multiple past cycles. The theoretical power consumption and the average measured power consumption are then weighted and fused according to a preset weight ratio (e.g., measured values ​​have a primary weight, and theoretical values ​​have a secondary weight), and the resulting value is used as the dynamic average power consumption for the current cycle. This mechanism ensures that the model is both theoretically instructive and can keenly follow the drift of actual working conditions.

[0035] To further improve the long-term stability of the model, this embodiment also introduces a self-updating mechanism for reference parameters. When the system detects that the deviation between the measured average power consumption over a long period and the theoretical average power consumption exceeds a preset tolerance ratio, it triggers a correction of the reference power consumption at the standard temperature. The correction logic adopts a weighted update method, that is, it uses the old reference power consumption and the measured power consumption after inverse normalization by the temperature coefficient to perform a weighted calculation to obtain the updated reference power consumption. In addition, the system also performs horizontal cluster comparison learning, classifying the current sensor into a sensor cluster with the same configuration and similar temperature environment. The system calculates the average power consumption of all sensors in the cluster as a reference benchmark. If the power consumption of the current sensor is significantly higher than the cluster average level, the system will use the cluster benchmark to forcibly calibrate the model parameters of that individual sensor, thereby eliminating the evaluation bias caused by individual hardware defects.

[0036] Finally, when calculating the remaining operating time, the system incorporates forward-looking future operating condition prediction logic. The system uses a time-series analysis algorithm to process historical temperature data. This algorithm comprehensively considers recent and long-term temperature data, giving greater weight to recent data and incorporating the prediction error from the previous period as a correction term, thereby predicting the average temperature over a preset future time period (e.g., the next week). Subsequently, the system adjusts the dynamic power consumption model based on this predicted temperature to calculate the expected average power consumption in the future. Finally, the system subtracts the accumulated charge consumption from the current actual available capacity to obtain the remaining battery power, divides this remaining power by the expected average power consumption in the future, and outputs the remaining battery operating time in days after unit conversion. This result can also be combined with a preset error upper limit (e.g., 8%) to output a prediction range including a confidence interval, providing users with a reference for conservative and optimistic estimates.

Claims

1. A method for adaptive evaluation of battery life for a wireless vibration sensor, the method comprising: The method includes: ​ Acquire the sensor's coulomb cumulative charge, ambient temperature, and configuration information including the acquisition interval and upload interval; The corresponding temperature decay factor is determined based on the ambient temperature, and the rated capacity of the battery is corrected in combination with the degree of battery aging to obtain the current actual usable capacity. Based on the configuration information, the reference theoretical power consumption is matched, and the actual power consumption calculated by the cumulative charge of the coulomb counter is used to dynamically correct the reference theoretical power consumption, thereby constructing a dynamic power consumption model that reflects the changes in operating conditions. The future ambient temperature is predicted based on historical temperature data, and the dynamic power consumption model is adjusted accordingly to obtain the future predicted power consumption. The remaining battery operating time is calculated by combining the current actual available capacity and the current remaining power.

2. The method of claim 1, wherein, The method divides the operating temperature range from -40℃ to 85℃ into several preset temperature intervals, determines the target interval to which the current ambient temperature belongs, and calculates the temperature decay factor using the linear correction coefficient corresponding to the target interval.

3. The method of claim 1 or 2, wherein, The method obtains the cumulative charge-discharge cycle count and the used time of the battery; The loss of active material caused by the cumulative number of charge-discharge cycles is calculated using the cyclic aging factor, and the self-discharge loss caused by the time of use is calculated using the calendar aging factor. The two are then combined to obtain the aging degradation factor. The rated capacity, temperature decay factor, and aging decay factor are multiplied to obtain the current actual usable capacity.

4. The method of claim 1, wherein, The method queries a preset configuration-power consumption mapping table based on the acquisition interval and the upload interval to obtain the theoretical reference power consumption of the current configuration at the standard temperature, and uses the current ambient temperature to perform temperature compensation on the theoretical reference power consumption to obtain the initial theoretical average power consumption.

5. The method of claim 1 or 4, wherein, The method determines whether the number of operation cycles of the sensor has reached a preset stability threshold; If the stable threshold is not reached, the initial theoretical average power consumption is directly determined as the dynamic power consumption of the current cycle. If a stable threshold has been reached, the actual average power consumption of multiple historical periods is calculated, and the initial theoretical average power consumption and the actual average power consumption are weighted and fused according to a preset weight. The fused result is determined as the dynamic power consumption of the current period.

6. The method of claim 1 or 4, wherein, The method divides the current sensors into a sensor cluster with the same configuration information and in the same temperature range; Calculate the average power consumption of all sensors in the sensor cluster as the cluster benchmark; If the current dynamic power consumption of the sensor exceeds the preset deviation range of the cluster benchmark, the sensor power consumption is determined to be abnormal, and the theoretical benchmark power consumption of the current sensor is corrected in reverse using the cluster benchmark.

7. The method for adaptive evaluation of battery life for wireless vibrational sensor of claim 1, wherein, The method obtains the actual working time before the battery replacement after detecting a battery replacement event; Calculate the assessment error between the actual working time and the estimated remaining working time before system replacement; If the evaluation error exceeds the preset tolerance threshold, the calculation parameters of the temperature decay factor or the theoretical reference power consumption of the configuration information will be adjusted according to the positive or negative direction of the error.

8. A wireless vibration sensor battery life adaptive evaluation system, the system is applied to the wireless vibration sensor battery life adaptive evaluation method of any one of claims 1 to 7, characterized in that, The system includes: The power input terminal of the coulomb counter module is electrically connected to the battery, the power output terminal of the coulomb counter module is electrically connected to the power management unit, and the data communication terminal of the coulomb counter module is connected to the microcontroller. The data output terminal of the battery temperature acquisition unit is connected to the microcontroller; The voltage output terminal of the power management unit is electrically connected to the microcontroller, the wireless communication unit, and the signal acquisition unit, respectively. The microcontroller is connected to the coulomb meter module and the battery temperature acquisition unit via a data bus. The microcontroller stores a sensor configuration information table containing the acquisition interval and upload interval.

9. A wireless vibrating sensor battery life self-adaptive evaluation system according to claim 8, wherein, The system also includes a signal acquisition unit and a data buffer unit; The power supply terminal of the signal acquisition unit is electrically connected to the voltage output terminal of the power management unit, and the signal output terminal of the signal acquisition unit is physically connected to the input interface of the microcontroller. The signal acquisition unit includes an accelerometer for sensing vibration signals. The data cache unit is electrically connected to the power management unit and the storage expansion interface of the microcontroller, respectively. The data cache unit is a non-volatile memory chip.

10. A wireless vibrating sensor battery life self-adapting evaluation system according to claim 8 or 9, characterized in that, The coulomb counter module is connected in series between the positive terminal of the battery and the input terminal of the power management unit; The coulomb counter module integrates a current sampling resistor and an analog-to-digital converter circuit. The current sampling resistor is connected in series in the power supply circuit, and the digital signal output pin of the analog-to-digital converter circuit is connected to the communication interface of the microcontroller via an I2C bus or an SPI bus.

Images