Electric quantity prediction method and system of backup battery, medium and product

Abstract

The invention discloses an electric quantity prediction method and system of a backup battery and a medium, and belongs to the related technical field of battery electric quantity prediction, and the method comprises the steps: setting a buffer module, carrying out the switching-on and switching-off simulation test of an instantaneous large-current pulse load working condition switch, and collecting voltage and current data reflecting the performance of the backup battery; calculating a plurality of features representing the dynamic response of the battery under the pulse load, and combining the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery to obtain a combined input vector; inputting the combined input vector into a preset opening and closing energy demand prediction model for analysis to obtain an available power consumption prediction index of the backup battery; and obtaining an electric quantity test report of the backup battery by comparing the available electric quantity prediction index with a preset protection electric quantity threshold value. According to the technical scheme of the invention, the available power of the backup battery can be accurately and reliably predicted under the real simulation switch opening and closing instantaneous large-current pulse load condition.

Description

Technical Field

[0001] This application relates to the technical field of battery power prediction, specifically to a method, system, and medium for predicting the power of a backup battery. Background Technology

[0002] As the core control and operation component of high-voltage switchgear, the reliable operation of the switch opening and closing mechanism directly affects the safe and stable operation of the power grid. In the event of fault tripping, sudden power outage, or low-voltage power supply failure, the backup battery becomes an important energy source to ensure the correct opening and closing of the switch. Since the backup battery is normally in a fully charged state of real-time charging, the actual usable power and voltage will change after the battery discharges, and may not be able to meet the daily functions of the terminal. When the actual usable power of the battery is insufficient, efficient operation cannot be performed. Therefore, it is necessary to detect the actual status of the backup power supply in the distribution automation terminal.

[0003] The existing methods for predicting backup battery capacity have the following drawbacks: Some newer batteries have built-in smart charging plugs, which actively increase the battery's internal resistance to a relatively high value when the battery is fully charged to prevent prolonged charging from affecting its lifespan. However, this affects the detection of the battery's internal resistance, and judging whether the battery can continue to operate by detecting the battery's internal resistance cannot accurately reflect the battery's dynamic response characteristics under instantaneous high current and short-term pulse load conditions, thus affecting the accuracy and reliability of battery capacity assessment tests.

[0004] Therefore, current technologies have the technical problem of being unable to accurately assess the dynamic response and actual available power of backup batteries under real-world simulation of high-current pulse load conditions during the opening and closing of circuit breakers. Summary of the Invention

[0005] This invention provides a method, system, medium, and product for predicting the power of a backup battery, which can solve the problem of accurately assessing the dynamic response and actual available power of a backup battery under the condition of instantaneous high current pulse load during the opening and closing of a simulated switch.

[0006] This invention provides a method for predicting the power capacity of a backup battery, comprising:

[0007] The control buffer module performs a switch opening and closing simulation test on the backup battery under instantaneous high current pulse load conditions, and collects the first voltage data and first current data of the backup battery under the current conditions; wherein, the buffer module is connected to the backup battery and is used to simulate the switch opening and closing instantaneous high current pulse load conditions of the backup battery.

[0008] Based on the first voltage data and the first current data, multiple features that can characterize the dynamic response of the backup battery under pulse load are calculated, and combined with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery to obtain a merged input vector.

[0009] The merged input vector is input into a preset opening and closing energy demand prediction model for analysis to obtain the available power prediction index of the backup battery.

[0010] Based on the available power prediction index, a power test report for the backup battery is generated.

[0011] This application utilizes a buffer module connected to the backup battery to conduct simulated switching on / off tests on the backup battery. This process collects voltage and current data reflecting instantaneous high-current pulse load conditions, providing a reliable data foundation for accurate and reliable prediction of the backup battery's available power under real-world conditions. By extracting the first voltage and first current data reflecting the backup battery's performance, multiple characteristics characterizing the battery's dynamic response under pulse loads are calculated from the original waveform data. These characteristics are then combined with equivalent instantaneous resistance and equivalent instantaneous energy to form a comprehensive characterization vector that fully describes the backup battery's state. This provides comprehensive data input for model prediction in subsequent steps, improving the scientific rigor of the prediction results. By merging the input vector and inputting it into the trained switching energy demand prediction model, the data reflecting the backup battery's performance is converted into predictive indicators, providing a decision-making basis for generating a backup battery power report. By analyzing the available power prediction indicators, the backup battery power report is obtained, ultimately achieving accurate prediction of the backup battery's power under instantaneous high-current pulse load conditions during switching on / off operations. Compared with existing technologies, this application achieves the technical effect of accurately and reliably predicting the available power of the backup battery under real operating conditions by setting a buffer module to simulate the instantaneous high current pulse load of the backup battery during switching on and off.

[0012] Furthermore, the control buffer module performs a switch-opening and closing simulation test on the backup battery under instantaneous high-current pulse load conditions, and collects the first voltage data and first current data of the backup battery under the current conditions, including:

[0013] The control buffer module includes: an energy storage unit, a bidirectional power converter, and a sampling unit;

[0014] The load simulation device of the bidirectional power converter receives a switch opening and closing simulation test command, drives the MCU controller of the bidirectional power converter to draw current from the backup battery, increase the voltage and then quickly charge the energy storage unit, or directly supply energy to the load; when the switch opening and closing test is completed, the MCU controller reduces the voltage of the energy stored in the energy storage unit and recharges it back to the backup battery; wherein, the load simulation device performs N consecutive switch opening and closing simulation tests, where N is a positive integer greater than or equal to 3;

[0015] The energy storage unit stores and releases the electrical energy released by the backup battery during the switch opening and closing test;

[0016] The sampling unit includes a voltage sensor and a current sensor. The voltage sensor is connected in parallel across the battery terminals, and the current sensor is connected in series in the main circuit flowing through the energy storage unit. Based on the N switching simulation tests, the corresponding N instantaneous voltage change waveforms and N instantaneous current change waveforms are obtained as the first voltage data and the first current data.

[0017] By incorporating a buffer module connected to the backup battery, this system, based on an energy storage unit, a bidirectional power converter, and a sampling unit, simulates the operation of a high-current pulse load during the switching on and off of a circuit breaker. The load simulation device within the bidirectional power converter enables simulated switching on and off tests under different parameter requirements, with a short, fixed time interval between each test to ensure battery voltage recovery. This effectively simulates the scenario of multiple consecutive operations of a switching device within a short period in real-world conditions. The MCU controller of the bidirectional power converter generates pulse-width modulation signals to perform boost discharge and buck charge operations on the backup battery. Combined with the energy storage unit, this allows for simulated switching on and off tests of the backup battery, reducing battery power consumption during the switching on and off tests themselves and preparing for subsequent tests. The energy storage unit enables rapid storage and release of power during simulated switching on and off. Furthermore, during boost discharge, the energy storage unit acts as an energy buffer, smoothing current surges and preventing extreme stress that could result from direct load application to the battery. This ensures the controllability and safety of the simulated switching on and off process and improves the accuracy of power prediction. By setting up a sampling unit that includes a voltage sensor and a current sensor, the instantaneous change waveform of the terminal voltage of the backup battery and the instantaneous change waveform of the current of the energy storage unit are acquired for subsequent extraction of voltage characteristic parameters and current characteristic parameters.

[0018] Further, generating a power test report for the backup battery based on the available power prediction index includes:

[0019] The power test report includes one or more of the following: battery health level, maximum number of circuit breakers / closers supported, and remaining available power.

[0020] The battery health level is obtained by predicting available power capacity.

[0021] The maximum number of circuit breakers that the remaining power of the backup battery can support is calculated by using the available power prediction index and the average energy consumption per operation.

[0022] The remaining available power is calculated using the available power forecast index and the rated capacity of the backup battery.

[0023] By setting the battery health level, the maximum number of circuit breakers that can be opened and closed, and the remaining available power as the power test report, the backup battery's predicted power level is comprehensively evaluated from multiple perspectives, including battery status and remaining power supply capacity, the number of complete circuit breakers that can be opened and closed, and the total energy that the battery can still release in the current state. This improves the accuracy of the backup battery power prediction results and provides a basis for determining whether to output a battery replacement prompt.

[0024] Furthermore, the method of obtaining the battery health level through the available power prediction index specifically includes:

[0025] The battery health level is obtained by comparing the available power prediction index with the first preset protection power threshold, the second preset protection power threshold, and the third preset protection power threshold under the corresponding test conditions.

[0026] Specifically, the protection power threshold after the initial trip is taken as the first preset protection power threshold, the protection power threshold after tripping and then closing is taken as the second preset protection power threshold, and the protection power threshold after closing and then tripping is taken as the third preset protection power threshold.

[0027] This method compares the available power prediction index step-by-step with multiple preset protection power thresholds—specifically, with the first, second, and third preset protection power thresholds under the corresponding test conditions—to determine whether the available power prediction index is greater than or equal to the first, second, or third preset protection power threshold, thereby obtaining a backup battery power test report. This improves the accuracy of the obtained backup battery power test report. The preset protection power thresholds are predefined numerical constants for the backup battery based on the minimum energy required for reliable operation of the high-voltage switchgear, safety margins, and other predefined values. By setting multiple preset protection power threshold ranges, the available power prediction problem is transformed into a mathematical problem, improving the understandability of the output results.

[0028] Furthermore, the preset energy demand prediction model for switching on and off is specifically as follows:

[0029] Based on multiple sets of backup battery test samples under different health conditions and labeled samples of actual available power indicators, the surrogate model is iteratively trained until the root mean square error meets the preset condition. The surrogate model of the current iteration is then output as the energy demand prediction model for switching on and off. In each iteration, the root mean square error between the predicted labeled samples of the surrogate model output and the labeled samples of the actual available power indicators is calculated. If the root mean square error is higher than the preset value, the surrogate model is optimized by adjusting the hyperparameters.

[0030] By collecting data from multiple backup batteries in different health states and conducting simulated switching operations, the known actual available power of each backup battery was accurately measured. This data was then used as training label samples for model training, improving the model's output accuracy and reliability. By merging input vector samples and training label samples to train the surrogate model, multiple decision trees were iteratively constructed to correct the prediction errors of the previous decision tree. Finally, a set of parameters was determined to minimize the overall difference between the model's predicted values ​​and the actual labels, serving as the surrogate model for predicting switching energy demand, thus enhancing the accuracy of the surrogate model's predictions. The trained surrogate model was evaluated by calculating the root mean square error (RMSE) between the surrogate model's predicted values ​​and the actual labels. For RMSE values ​​exceeding a preset threshold, the model's hyperparameters were adjusted iteratively until the threshold was met, improving the reliability of the output switching energy demand prediction model.

[0031] Furthermore, the step of calculating multiple features that characterize the dynamic response of the backup battery under pulsed load based on the first voltage data and the first current data specifically includes:

[0032] Based on the Nth instantaneous voltage change waveform and the Nth instantaneous current change waveform in the first voltage data and the first current data, extract the corresponding N sets of voltage characteristic parameters and N sets of current characteristic parameters.

[0033] Based on the N sets of voltage characteristic parameters and the N sets of current characteristic parameters, the voltage and current data reflecting the performance of the backup battery are recalculated.

[0034] By preprocessing the N-order instantaneous voltage and current change waveforms separately, using a moving average filter to denoise both waveforms, calculating the rate of change between adjacent data points, identifying and removing abnormal data points, and then correcting them using linear interpolation based on the normal data points before and after the abnormal data points, the accuracy of the data and the subsequent results are improved. Analyzing the preprocessed data extracts voltage and current characteristic parameters, providing fundamental data reflecting the backup battery's performance. By calculating the average values ​​of the N sets of voltage and current characteristic parameters, the resilience and accuracy of the final predicted battery capacity under continuous discharge conditions are improved.

[0035] Furthermore, the calculation of multiple features characterizing the dynamic response of the backup battery under pulsed load, combined with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery, yields a merged input vector, specifically:

[0036] The voltage characteristic parameters include: voltage drop magnitude, recovery time constant, and dynamic internal resistance parameter;

[0037] The voltage drop amplitude is calculated by comparing the stable voltage before the pulse load begins with the lowest voltage during the load period.

[0038] The voltage recovery time constant is calculated by taking the time required for the battery voltage to recover from the drop value to the preset stable value.

[0039] The dynamic internal resistance parameters are calculated by the instantaneous changes in voltage and current when a current pulse is applied.

[0040] The current characteristic parameters include: energy absorption peak value, current decay slope, and energy release time constant;

[0041] The peak current / peak energy absorption ratio is calculated by the maximum absolute value of the current waveform during the pulse.

[0042] The current decay slope is calculated by linearly fitting the slope of the falling edge of the current pulse.

[0043] The energy release time constant is calculated by measuring the duration of the current pulse.

[0044] The equivalent instantaneous resistance is calculated by synchronously acquiring voltage and current values ​​at a specific moment during the pulse load.

[0045] The equivalent instantaneous energy is calculated by using the total energy released during a single pulse load.

[0046] The voltage characteristic parameters, current characteristic parameters, equivalent instantaneous resistance, and equivalent instantaneous energy are combined to obtain a combined input vector.

[0047] By calculating multiple characteristics that characterize the dynamic response of the backup battery under instantaneous high-current pulse load conditions, and describing it from multiple aspects such as current parameters, voltage parameters, equivalent instantaneous resistance, and equivalent instantaneous energy, the correlation between characteristic parameters and the richness of electrical data are increased, thereby improving the accuracy of the output results of the opening and closing energy demand prediction model. By merging data points at the same time, complete electrical states at different times are obtained, providing a data decision basis for subsequent opening and closing energy demand prediction model predictions.

[0048] This invention provides a backup battery power prediction system, comprising: a simulated load unit, a characteristic parameter output unit, a demand prediction and analysis unit, and a test report acquisition unit.

[0049] The simulated load unit is used to control the buffer module to perform a simulated test of the switching on and off of the backup battery under the instantaneous high current pulse load condition, and to collect the first voltage data and the first current data of the backup battery under the current condition; wherein, the buffer module is connected to the backup battery and is used to simulate the working condition of the backup battery under the instantaneous high current pulse load condition of the switching on and off.

[0050] The feature parameter output unit is used to calculate multiple features that can characterize the dynamic response of the backup battery under pulse load based on the first voltage data and the first current data, and combine them with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery to obtain a merged input vector.

[0051] The demand forecasting and analysis unit is used to input the merged input vector into a preset opening and closing energy demand forecasting model for analysis, and to obtain the available power forecast index of the backup battery.

[0052] The test report acquisition unit is used to generate a power test report for the backup battery based on the available power prediction index.

[0053] Another embodiment of this application provides a computer-readable storage medium storing a computer program product thereon, which, when executed by a processor, implements the steps of the backup battery power prediction method of the present invention.

[0054] Another embodiment of this application also provides a computer program product, including a computer program or instructions, which, when executed by a communication device, implement the steps of the backup battery power prediction method of the present invention. Attached Figure Description

[0055] To more clearly illustrate the technical solutions of this application, the accompanying drawings used in the embodiments will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0056] Figure 1 This is a schematic flowchart of a backup battery power prediction method provided in an embodiment of this application.

[0057] Figure 2 This is a schematic diagram of a backup battery power prediction system provided in an embodiment of this application.

[0058] Explanation of reference numerals in the attached figures: Simulated load unit 501, characteristic parameter output unit 502, demand forecasting and analysis unit 503, test report acquisition unit 504. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. It should be understood that the following implementation process is merely illustrative and is intended to ensure that those skilled in the art can reproduce this application based on it. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0060] In the following description, references to "some embodiments" describe a subset of all possible embodiments. However, it is understood that "some embodiments" can be the same or different subsets of all possible embodiments and can be combined with each other without conflict. The terms "first" and "second" are used merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only.

[0061] Example 1

[0062] See Figure 1 To address the problem of backup battery power prediction under instantaneous high-current pulse load conditions in the prior art, this application provides a backup battery power prediction method according to an embodiment, including steps S1 to S4, the specific steps of which are as follows:

[0063] S1, control buffer module, performs switch opening and closing simulation test on backup battery under instantaneous high current pulse load condition, and collects the first voltage data and first current data of backup battery under the current condition.

[0064] The buffer module is connected to the backup battery and is used to simulate the working conditions of the backup battery under instantaneous high current pulse load during the switching on and off.

[0065] Furthermore, the control buffer module performs a switch-opening and closing simulation test on the backup battery under a transient high-current pulse load condition, and collects the first voltage data and first current data of the backup battery under the current condition. The control buffer module includes an energy storage unit, a bidirectional power converter, and a sampling unit. The load simulation device of the bidirectional power converter receives the switch-opening and closing simulation test command, drives the MCU controller of the bidirectional power converter to draw current from the backup battery, increase the voltage, and then rapidly charge the energy storage unit, or directly supply energy to the load. When the switch-opening and closing test is completed, the MCU controller... The stored electrical energy in the energy storage unit is stepped down and then recharged back to the backup battery; wherein, the load simulation device performs N consecutive switching on / off simulation tests, where N is a positive integer greater than or equal to 3; the energy storage unit stores and releases the electrical energy released by the backup battery during the switching on / off tests; the sampling unit includes a voltage sensor and a current sensor, wherein the voltage sensor is connected in parallel across the battery terminals, and the current sensor is connected in series in the main circuit flowing through the energy storage unit; based on the N switching on / off simulation tests, the corresponding N instantaneous voltage change waveforms and N instantaneous current change waveforms are obtained as first voltage data and first current data.

[0066] In one embodiment, the control buffer module performs a switch-opening and closing simulation test on the backup battery under a transient high-current pulse load condition, and collects the first voltage data and first current data of the backup battery under the current condition, including:

[0067] Preferably, the buffer module is a hardware module whose input / output terminals are connected to the positive and negative terminals of the backup battery via cables. The buffer module is composed of three sub-modules integrated in the circuit, including an energy storage unit, a bidirectional power converter, and a sampling unit.

[0068] The energy storage unit may be a supercapacitor / electrolytic capacitor or an inductor, capable of rapidly storing and releasing electrical energy. The bidirectional power converter is a power electronic circuit, typically composed of a topology consisting of fully controlled switching devices such as MOSFETs, like an H-bridge, and equipped with a drive circuit and a pulse width modulation control chip. By controlling the switching devices, it enables bidirectional current flow between the backup battery and the energy storage unit. Specifically, the bidirectional power converter controls the energy exchange between the energy storage unit and the backup battery during switching operations, providing buffer energy at the moment of switching to reduce the direct inrush current to the backup battery and improve the accuracy of power prediction. The sampling unit refers to the specific sensor circuit, including a voltage sensor and a current sensor. The voltage sensor is connected in parallel across the battery terminals to measure the battery terminal voltage, and the current sensor is connected in series in the main circuit to measure the current flowing through the energy storage unit.

[0069] During simulated opening or closing of the circuit breaker, the bidirectional power converter is controlled in boost mode, drawing current from the backup battery, increasing its voltage, and then rapidly charging the energy storage unit or directly supplying energy to the load. In this state, the battery is discharging. When the simulation ends or testing is required, the bidirectional power converter is controlled in buck mode, reducing the voltage of the energy storage unit and recharging it back to the backup battery. In this state, the battery is charging. When the load simulation device performs switch opening and closing simulation tests on the backup battery based on the bidirectional power converter, the sampling unit synchronously acquires the instantaneous waveform changes of the backup battery's terminal voltage and the instantaneous waveform changes of the energy storage unit's current.

[0070] In this circuit breaker simulation test, the energy path is dominated by a bidirectional power converter. The energy released by the backup battery is transmitted through the power circuit of the bidirectional power converter. Then, a controlled, instantaneous high-current discharge is directly applied to the backup battery, subjecting it to the same electrical stress as in actual operating conditions. Due to the battery's internal resistance, its terminal voltage drops when the instantaneous high current is extracted, maintaining a low voltage level during current stabilization. The voltage recovers after the current is removed. The voltage sensor continuously measures the voltage value across the battery at a high sampling rate, forming a voltage instantaneous change waveform. The current sensor continuously measures the current flowing into / out of the energy storage unit at the same synchronous high sampling rate, forming a current instantaneous change waveform. Finally, voltage and current data reflecting the battery's performance under real pulse load are obtained. In this way, by setting up a buffer module, the instantaneous high-current pulse load condition simulation of the backup battery's power capacity during circuit breaker operation is achieved.

[0071] Preferably, the load simulation device is the controller of the bidirectional power converter, typically a microcontroller unit (MCU). It receives switch opening and closing simulation test commands through an interface, which may include test parameters such as pulse width, target current value, and number of tests. Based on the switch opening and closing simulation test commands, it drives the MCU controller of the bidirectional power converter to perform boost discharge. Specifically, the MCU generates a pulse width modulation signal according to the received switch opening and closing simulation test commands to drive the power switch in the bidirectional power converter, putting it in boost conversion mode. That is, the current flows out from the backup battery, passes through the bidirectional power converter, and after the voltage is boosted, it quickly charges the energy storage unit or directly supplies power to the simulated load. At this time, the battery is in a high-current discharge state.

[0072] Preferably, the MCU controller buffers the backup battery based on the energy storage unit. That is, during the boost discharge process, the energy storage unit receives energy from the battery as an instantaneous load, and at the same time acts as an energy buffer to smooth the current impact, avoiding the extreme stress that may be caused by the load being directly applied to the battery, and ensuring the controllability and safety of the switch opening and closing simulation test process.

[0073] Preferably, when the switch opening and closing test is completed, the MCU controller is driven to perform buck charging. That is, the MCU changes the duty cycle and phase of its output PWM signal, and the bidirectional power converter switches from boost conversion mode to buck conversion mode. Current flows out from the energy storage unit, and after the voltage is reduced by the bidirectional power converter, it is returned to the backup battery. That is, the backup battery is in a charging state, which reduces the power consumption of the battery by the switch opening and closing test itself and prepares for the next test. The voltage of the energy storage unit is restored to the initial state, waiting for the next switch opening and closing simulation test command to perform the switch opening and closing simulation test.

[0074] The load simulation device's controller is programmed to automatically and continuously execute N pulse discharge tests identical to a single switch opening and closing simulation test, with a short, fixed time interval between each discharge to ensure battery voltage recovery. A sampling unit synchronously acquires the instantaneous waveforms of the terminal voltage and the instantaneous current changes of the energy storage unit corresponding to the 1st to Nth instantaneous discharge processes. In this way, N instantaneous discharges are used to simulate N opening and closing operations, i.e., simulating multiple consecutive operations of a switching device in a short period of time in reality. N is a positive integer greater than or equal to 3, ensuring that the test can effectively simulate continuous operation and that the obtained data is realistic.

[0075] The sampling times of voltage and current data are strictly synchronized by the same controller clock source to ensure that each voltage data point has a corresponding current data point. In this way, the data samples can be aligned through timestamps, providing the necessary conditions for subsequent data merging.

[0076] S2, based on the first voltage data and the first current data, calculate multiple features that can characterize the dynamic response of the backup battery under pulse load, and combine them with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery to obtain a merged input vector.

[0077] Furthermore, the step of calculating multiple features that characterize the dynamic response of the backup battery under pulsed load based on the first voltage data and the first current data specifically involves: extracting N sets of voltage feature parameters and N sets of current feature parameters based on the Nth instantaneous voltage change waveform and the Nth instantaneous current change waveform in the first voltage data and the first current data; and recalculating the voltage data and current data reflecting the performance of the backup battery based on the N sets of voltage feature parameters and the N sets of current feature parameters.

[0078] Further, the calculation of multiple features characterizing the dynamic response of the backup battery under pulsed load, combined with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery, is merged to obtain a merged input vector. Specifically, the voltage characteristic parameters include: voltage drop amplitude, recovery time constant, and dynamic internal resistance parameter; the voltage drop amplitude is calculated using the stable voltage before the pulsed load begins and the lowest voltage during the load; the voltage recovery time constant is calculated using the time required for the battery voltage to recover from the drop value to the preset stable value; the dynamic internal resistance parameter is calculated using the change in voltage and the change in current at the instant the current pulse is applied; the current characteristic parameters include... The parameters include: peak energy absorption, current decay slope, and energy release time constant; the peak current / peak energy absorption ratio is calculated using the maximum absolute value of the current waveform during the pulse; the current decay slope is calculated using the linear fitting slope of the falling edge of the current pulse; the energy release time constant is calculated using the duration width of the current pulse; the equivalent instantaneous resistance is calculated using the voltage and current values ​​synchronously acquired at a specific moment during the pulse load; the equivalent instantaneous energy is calculated using the total energy released during a single pulse load; and the voltage characteristic parameters, current characteristic parameters, equivalent instantaneous resistance, and equivalent instantaneous energy are combined to obtain a combined input vector.

[0079] In one embodiment, based on the first voltage data and the first current data, multiple features characterizing the dynamic response of the backup battery under pulsed load are calculated, and combined with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery to obtain a merged input vector, including steps S201 to S202, each step of which is as follows:

[0080] S201, preprocess the first voltage data and the first current data.

[0081] Specifically, for voltage and current data points occurring simultaneously, a moving average filter is used to denoise the instantaneous voltage and current waveforms. This involves sliding a fixed-width window across the data and calculating the arithmetic mean of the data within the window as the output for that point, thus eliminating or reducing high-frequency random fluctuations mixed in with the real signal. Next, the rate of change between adjacent data points is calculated to identify data points caused by strong instantaneous interference, sampling errors, or transmission errors, and these are marked as outliers. These outlier data points are then removed, and reasonable replacement values ​​are calculated using linear interpolation with the values ​​of the normal data points before and after them for correction. This preprocessing of the first voltage and current data enhances the accuracy of the data and improves the accuracy of subsequent prediction results.

[0082] S202 calculates the voltage characteristic parameters, current characteristic parameters, equivalent instantaneous resistance, and equivalent instantaneous energy based on the preprocessed data to obtain the merged input vector.

[0083] Specifically, the instantaneous waveforms of the Nth-order terminal voltage change and the Nth-order instantaneous current change are processed and analyzed to extract and output N sets of voltage characteristic parameters, N sets of current characteristic parameters, equivalent instantaneous resistance, and equivalent instantaneous energy corresponding to the Nth-order terminal voltage change and the Nth-order instantaneous current change waveforms, respectively. Then, the average values ​​of the N sets of voltage characteristic parameters, N sets of current characteristic parameters, equivalent instantaneous resistance, and equivalent instantaneous energy are calculated and a merged input vector is constructed. This vector is then input into the opening and closing energy demand prediction model for analysis. Finally, a final power prediction value that comprehensively reflects the performance of the backup battery under continuous discharge conditions and has stronger anti-fluctuation ability is recalculated and used as the available power prediction index of the backup battery.

[0084] Preferably, the voltage characteristic parameters include voltage drop amplitude, recovery time constant, and dynamic internal resistance parameter; the current characteristic parameters include energy absorption peak value, current decay slope, and energy release time constant. Specifically, the voltage drop amplitude is obtained by calculating the difference between the stable voltage before the pulse load begins and the lowest voltage point during the load, directly reflecting the battery's transient load capacity; the voltage recovery time constant is obtained by determining the time required for the battery voltage to recover from the drop value to a stable value of a specific percentage (e.g., 95%) at the end of the pulse load, reflecting the battery's polarization recovery speed; the dynamic internal resistance parameter is estimated by calculating the ratio of the voltage change to the current change at the instant the current pulse is applied; the maximum absolute value of the current waveform during the pulse is identified to obtain the current peak value / energy absorption peak value; the linear fitting slope of the current pulse falling edge is calculated to obtain the current decay slope; the duration of the entire current pulse is quantified, and the energy release time constant is obtained through exponential fitting.

[0085] Specifically, the equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery under the energy storage unit are identified. At a specific moment during a pulse load, based on synchronously acquired voltage and current values, the equivalent instantaneous resistance is calculated using the instantaneous form of Ohm's law, reflecting the total load impedance at that operating point. Then, the total energy released during a single pulse load is integrated to identify the equivalent instantaneous energy, representing the actual energy consumed to complete one opening or closing operation. Next, voltage characteristic parameters, current characteristic parameters, equivalent instantaneous resistance, and equivalent instantaneous energy are merged, arranged in a predefined order to form a merged input vector, comprehensively characterizing the battery's dynamic response and energy consumption. By calculating multiple features that characterize the backup battery's dynamic response under instantaneous high-current pulse load conditions, the richness of electrical data is increased, improving the accuracy of the opening and closing energy demand prediction model output. By merging data points at the same time, complete electrical states at different times are obtained, providing a data decision-making basis for subsequent opening and closing energy demand prediction model predictions.

[0086] S3, input the merged input vector into the preset opening and closing energy demand prediction model for analysis, and obtain the available power prediction index of the backup battery.

[0087] Furthermore, the preset energy demand prediction model for switching on and off is specifically as follows:

[0088] Based on multiple sets of backup battery test samples under different health conditions and labeled samples of actual available power indicators, the surrogate model is iteratively trained until the root mean square error meets the preset condition. The surrogate model of the current iteration is then output as the energy demand prediction model for switching on and off. In each iteration, the root mean square error between the predicted labeled samples of the surrogate model output and the labeled samples of the actual available power indicators is calculated. If the root mean square error is higher than the preset value, the surrogate model is optimized by adjusting the hyperparameters.

[0089] In one embodiment, the step of inputting the merged input vector into a preset opening and closing energy demand prediction model for analysis to obtain the available power prediction index of the backup battery includes steps S301 to S302, each step of which is as follows:

[0090] S301, training the energy demand prediction model for switching on and off.

[0091] Preferably, multiple backup batteries with different aging levels and capacity decay states are collected and subjected to switch-on / off simulation tests. Multiple sets of instantaneous voltage and current change waveforms are output as multiple backup battery test samples. Then, equivalent instantaneous parameters are identified based on the backup battery test samples and merged to generate a merged input vector sample. Using standard capacity verification methods, such as constant current discharge testing, the known actual usable capacity of the backup battery is accurately measured and used as the corresponding training label sample. Next, a prediction model is constructed based on the XGBoost regression model. The prediction model is trained using the merged input vector sample and the training label sample. Multiple decision trees are iteratively constructed to correct the previous... The prediction error of each decision tree is used to determine a set of parameters that minimizes the overall difference between the model's predicted values ​​and the true labels, thus obtaining the energy demand prediction model for circuit breaker opening and closing. Then, the root mean square error (RMSE) of the trained energy demand prediction model is evaluated, i.e., the root mean square error between the model's predicted values ​​and the true labels is calculated to quantify the average error of the model's predictions. The energy demand prediction model is then retrained based on the RMS error. If the RMS error exceeds a preset RMS error threshold constraint, the energy demand prediction model is considered inaccurate, and the model's hyperparameters are adjusted to seek better model parameters until the RMS error threshold constraint is met, at which point training terminates, and the trained energy demand prediction model is output and downloaded. By collecting multiple sets of backup batteries in different health states as samples for model training and iterating multiple times to determine a surrogate model with parameters that minimize the overall difference between the model's predicted values ​​and the true labels, the accuracy and reliability of the energy demand prediction model for circuit breaker opening and closing are improved.

[0092] S302, input the merged input vector into the opening and closing energy demand prediction model to obtain the available power prediction index of the backup battery.

[0093] S4. Generate a power test report for the backup battery based on the available power prediction index.

[0094] Further, generating a power test report for the backup battery based on the available power prediction index includes: the power test report includes one or more of the following: battery health level, maximum number of circuit breakers / closers supported, and remaining available power; obtaining the battery health level through the available power prediction index; calculating the maximum number of circuit breakers / closers supported by the remaining power of the backup battery through the available power prediction index and the average energy consumption per operation; and calculating the remaining available power through the available power prediction index and the rated capacity of the backup battery.

[0095] Furthermore, the step of obtaining the battery health level through the available power prediction index specifically involves: comparing the available power prediction index with the corresponding first preset protection power threshold, second preset protection power threshold, and third preset protection power threshold under the test conditions to obtain the battery health level; wherein, the protection power threshold after the initial trip is taken as the first preset protection power threshold, the protection power threshold after tripping and then closing is taken as the second preset protection power threshold, and the protection power threshold after closing and then tripping is taken as the third preset protection power threshold.

[0096] In one embodiment, generating a power test report for the backup battery based on the available power prediction index includes:

[0097] The preset protection power thresholds include at least a first preset protection power threshold, a second preset protection power threshold, and a third preset protection power threshold. The available power prediction index of the backup battery is re-obtained and compared with the first, second, and third preset protection power thresholds under the corresponding test conditions to obtain a power test report for the backup battery. The power test report includes the battery health level, the maximum number of circuit breakers / closers supported, and the remaining available power percentage. If any one of the battery health level, the maximum number of circuit breakers / closers supported, or the remaining available power percentage fails to meet the corresponding threshold, a battery replacement prompt is output.

[0098] The preset protection power threshold is a predefined numerical constant of the backup battery based on the minimum energy required for reliable operation of the high-voltage switchgear and the safety margin. The preset protection power threshold includes at least a first preset protection power threshold, a second preset protection power threshold, and a third preset protection power threshold, defining the minimum energy reserve required for the high-voltage switchgear to complete different operations. The third preset protection power threshold > the second preset protection power threshold > the first preset protection power threshold. The first preset protection power threshold is the protection power threshold after the initial trip, that is, the minimum energy required to ensure that the backup battery can complete at least one trip operation. The second preset protection power threshold is the protection power threshold after tripping and then closing, that is, to ensure that after completing one trip, the remaining energy of the battery is still sufficient to complete one closing operation. The third preset protection power threshold is the protection power threshold after closing and then tripping, that is, to ensure that after completing the tripping-closing operation, the remaining energy of the battery is still sufficient to complete another tripping operation.

[0099] The process involves comparing the newly obtained available power prediction index with preset protection power thresholds step by step. Specifically, it compares the index with the first, second, and third preset protection power thresholds under the corresponding test conditions to determine whether the available power prediction index is greater than or equal to these thresholds. This results in a backup battery power test report, which includes the battery health level, the maximum number of circuit breakers / closes that can be supported, and the remaining available power percentage. Specifically, the battery health level is a classification label based on the level of the available power prediction index relative to the preset protection power thresholds. If the available power prediction index > the third preset protection power threshold, the battery health level is excellent; if the second preset protection power threshold < the available power prediction index < the third preset protection power threshold, the battery health level is good; if the first preset protection power threshold < the available power prediction index < the second preset protection power threshold, the battery health level is warning; and if the available power prediction index < the first preset protection power threshold, the battery health level is dangerous. By comparing the available power prediction index with multiple preset protection power thresholds step by step, the available power prediction problem is transformed into a mathematical problem, improving the understandability of the output results.

[0100] The maximum number of circuit breakers that can be opened and closed is calculated based on the available power prediction index and the average energy consumption per operation, representing the number of times the backup battery can perform a complete operation. The remaining available power percentage is calculated based on the available power prediction index and the backup battery's rated capacity. If any of the battery health level, maximum number of circuit breakers that can be opened and closed, or remaining available power percentage in the power test report does not meet the corresponding threshold (where the corresponding threshold may be a dangerous health level, a maximum number of operations < 1, or a remaining power percentage < 20%), a battery replacement prompt will be automatically generated and output, such as "Warning, battery power is insufficient, unable to guarantee one circuit breaker operation, please replace immediately." This comprehensive evaluation of the backup battery's power capacity from three aspects—battery health level, maximum number of circuit breakers that can be opened and closed, and remaining available power percentage—improves the scientific rigor and practicality of the final judgment.

[0101] Another embodiment of this application provides a backup battery power prediction system, including: a simulated load unit 501, a characteristic parameter output unit 502, a demand prediction analysis unit 503, and a test report acquisition unit 504.

[0102] The simulated load unit 501 is used to control the buffer module to perform a switch opening and closing simulation test on the backup battery under the instantaneous high current pulse load condition, and to collect the first voltage data and the first current data of the backup battery under the current condition; wherein, the buffer module is connected to the backup battery and is used to simulate the working condition of the backup battery under the instantaneous high current pulse load condition of the switch opening and closing.

[0103] The feature parameter output unit 502 is used to calculate multiple features that can characterize the dynamic response of the backup battery under pulse load based on the first voltage data and the first current data, and combine them with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery to obtain a merged input vector.

[0104] The demand forecasting and analysis unit 503 is used to input the merged input vector into a preset opening and closing energy demand forecasting model for analysis, and obtain the available power forecast index of the backup battery.

[0105] The test report acquisition unit 504 is used to generate a power test report for the backup battery based on the available power prediction index.

[0106] It is understood that the above system embodiments correspond to the method embodiments of this application, and can implement the backup battery power prediction method provided by any of the above method embodiments of this application.

[0107] It should be noted that the system embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the system embodiments provided in this application, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0108] For ease of description and brevity, the system embodiments of this application include all the implementation methods in the above-described backup battery power prediction method embodiments, and will not be repeated here.

[0109] Based on the above embodiments of the backup battery power prediction method, another embodiment of this application provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the compression method of the electrical equipment state prediction network of any embodiment of this application.

[0110] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete this application. The one or more module units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0111] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0112] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0113] Based on the above-described method embodiments, another embodiment of this application provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the backup battery power prediction method described in any of the above-described method embodiments of this application.

[0114] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

Claims

1. A method for predicting the power capacity of a backup battery, characterized in that, The method includes: The control buffer module performs a switch opening and closing simulation test on the backup battery under instantaneous high current pulse load conditions, and collects the first voltage data and first current data of the backup battery under the current conditions; wherein, the buffer module is connected to the backup battery and is used to simulate the switch opening and closing instantaneous high current pulse load conditions of the backup battery. Based on the first voltage data and the first current data, multiple features that can characterize the dynamic response of the backup battery under pulse load are calculated, and combined with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery to obtain a merged input vector. The merged input vector is input into a preset opening and closing energy demand prediction model for analysis to obtain the available power prediction index of the backup battery. Based on the available power prediction index, a power test report for the backup battery is generated.

2. The backup battery power prediction method as described in claim 1, characterized in that, The control buffer module performs a simulated switch opening and closing test on the backup battery under instantaneous high-current pulse load conditions, and collects the first voltage data and first current data of the backup battery under the current conditions, including: The control buffer module includes: an energy storage unit, a bidirectional power converter, and a sampling unit; The load simulation device of the bidirectional power converter receives a switch opening and closing simulation test command, drives the MCU controller of the bidirectional power converter to draw current from the backup battery, increase the voltage and then quickly charge the energy storage unit, or directly supply energy to the load; when the switch opening and closing test is completed, the MCU controller reduces the voltage of the energy stored in the energy storage unit and recharges it back to the backup battery; wherein, the load simulation device performs N consecutive switch opening and closing simulation tests, where N is a positive integer greater than or equal to 3; The energy storage unit stores and releases the electrical energy released by the backup battery during the switch opening and closing test; The sampling unit includes a voltage sensor and a current sensor. The voltage sensor is connected in parallel across the battery terminals, and the current sensor is connected in series in the main circuit flowing through the energy storage unit. Based on the N switching simulation tests, the corresponding N instantaneous voltage change waveforms and N instantaneous current change waveforms are obtained as the first voltage data and the first current data.

3. The backup battery power prediction method as described in claim 1, characterized in that, The step of generating a power test report for the backup battery based on the available power prediction index includes: The power test report includes one or more of the following: battery health level, maximum number of circuit breakers / closers supported, and remaining available power. The battery health level is obtained by predicting available power capacity. The maximum number of circuit breakers that the remaining power of the backup battery can support is calculated by using the available power prediction index and the average energy consumption per operation. The remaining available power is calculated using the available power forecast index and the rated capacity of the backup battery.

4. The backup battery power prediction method as described in claim 3, characterized in that, The battery health level is obtained through the available power prediction index, specifically as follows: The battery health level is obtained by comparing the available power prediction index with the first preset protection power threshold, the second preset protection power threshold, and the third preset protection power threshold under the corresponding test conditions. Specifically, the protection power threshold after the initial trip is taken as the first preset protection power threshold, the protection power threshold after tripping and then closing is taken as the second preset protection power threshold, and the protection power threshold after closing and then tripping is taken as the third preset protection power threshold.

5. The backup battery power prediction method as described in claim 1, characterized in that, The preset energy demand prediction model for switching on and off is specifically as follows: Based on multiple sets of backup battery test samples under different health conditions and labeled samples of actual available power indicators, the surrogate model is iteratively trained until the root mean square error meets the preset condition. The surrogate model of the current iteration is then output as the energy demand prediction model for switching on and off. In each iteration, the root mean square error between the predicted labeled samples of the surrogate model output and the labeled samples of the actual available power indicators is calculated. If the root mean square error is higher than the preset value, the surrogate model is optimized by adjusting the hyperparameters.

6. The backup battery power prediction method as described in claim 1, characterized in that, The step of calculating multiple features that characterize the dynamic response of the backup battery under pulse load based on the first voltage data and the first current data specifically includes: Based on the Nth instantaneous voltage change waveform and the Nth instantaneous current change waveform in the first voltage data and the first current data, extract the corresponding N sets of voltage characteristic parameters and N sets of current characteristic parameters. Based on the N sets of voltage characteristic parameters and the N sets of current characteristic parameters, the voltage and current data reflecting the performance of the backup battery are recalculated.

7. The backup battery power prediction method as described in claim 1, characterized in that, The calculation of multiple features characterizing the dynamic response of the backup battery under pulsed load, combined with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery, yields a merged input vector, specifically: The voltage characteristic parameters include: voltage drop magnitude, recovery time constant, and dynamic internal resistance parameter; The voltage drop amplitude is calculated by comparing the stable voltage before the pulse load begins with the lowest voltage during the load period. The voltage recovery time constant is calculated by taking the time required for the battery voltage to recover from the drop value to the preset stable value. The dynamic internal resistance parameters are calculated by the instantaneous changes in voltage and current when a current pulse is applied. The current characteristic parameters include: energy absorption peak value, current decay slope, and energy release time constant; The peak current / peak energy absorption ratio is calculated by the maximum absolute value of the current waveform during the pulse. The current decay slope is calculated by linearly fitting the slope of the falling edge of the current pulse. The energy release time constant is calculated by measuring the duration of the current pulse. The equivalent instantaneous resistance is calculated by synchronously acquiring voltage and current values ​​at a specific moment during the pulse load. The equivalent instantaneous energy is calculated by using the total energy released during a single pulse load. The voltage characteristic parameters, current characteristic parameters, equivalent instantaneous resistance, and equivalent instantaneous energy are combined to obtain a combined input vector.

8. A backup battery power prediction system, characterized in that, The system is used to implement the backup battery power prediction method according to any one of claims 1 to 7, the system comprising: The simulated load unit is used to control the buffer module to perform a simulated test of the switching on and off of the backup battery under the instantaneous high current pulse load condition, and to collect the first voltage data and the first current data of the backup battery under the current condition; wherein, the buffer module is connected to the backup battery and is used to simulate the working condition of the backup battery under the instantaneous high current pulse load condition of the switching on and off. The feature parameter output unit is used to calculate multiple features that can characterize the dynamic response of the backup battery under pulse load based on the first voltage data and the first current data, and combine them with the identified equivalent instantaneous resistance and equivalent instantaneous energy of the backup battery to obtain a merged input vector. The demand forecasting and analysis unit is used to input the merged input vector into a preset opening and closing energy demand forecasting model for analysis, and to obtain the available power forecast index of the backup battery. The test report acquisition unit is used to generate a power test report for the backup battery based on the available power prediction index.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program or instructions, which, when executed by a communication device, implement the backup battery power prediction method as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by the communication device, the backup battery power prediction method as described in any one of claims 1 to 7 is implemented.

Images