A commercial energy storage battery management system and method

Abstract

The application relates to the technical field of industrial control systems and discloses an industrial and commercial energy storage battery management system and method, which comprises the following steps: superimposing a micro-disturbance signal on a battery main loop switch device, collecting a voltage response sequence in real time and calculating an approximate entropy value, determining a battery operation mode according to entropy value threshold comparison and executing a corresponding control strategy. The application directly senses the internal electrochemical stability of the battery through the order degree analysis of the voltage response sequence, realizes model-free state recognition, simultaneously constructs a resource self-adaptive mechanism of macroscopic monitoring triggering microscopic calculation, verifies the electrochemical and thermodynamic signals under complex working conditions to improve the decision reliability, and forms a dynamic control system which takes into account the real-time performance and safety.

Description

TECHNICAL FIELD

[0001] The application relates to an industrial and commercial energy storage battery management system and belongs to the technical field of industrial control systems. BACKGROUND

[0002] The current industrial and commercial energy storage system generally adopts a state monitoring strategy based on an equivalent circuit model. The internal state is indirectly inferred by collecting battery macroscopic parameters through multiple sensors and inputting an algorithm model. It is generally found in the industry that when the system faces high dynamic scenarios such as grid frequency modulation or load mutation, the model updating speed is difficult to match the instantaneous changes of the internal electrochemical state of the battery. Notably, this dynamic response lag not only causes the control instruction to be mismatched with the actual carrying capacity of the battery, but also accumulates the risk of overcharging and overdischarging.

[0003] Taking the typical scenario of mill start-stop in an industrial park as an example: when the energy storage battery experiences high-rate charging and then immediately performs a large-current discharging task, the traditional system can determine that the power is sufficient through the state of charge model, but it cannot perceive the electrode polarization effect caused by the charging process. As a result, the electrode voltage drops suddenly at the moment of discharging, triggering protection shutdown and causing production interruption. More deeply, the existing technical architecture has three inherent defects: first, the static model has a blind area in recognizing the transient polarization response of the battery; second, the decision delay and algorithm bottleneck caused by multi-source data fusion; and third, the one-way control logic lacks a verification mechanism for the authenticity of the internal state.

[0004] Although the industry has tried to improve accuracy by increasing temperature monitoring points or strengthening algorithm complexity, it has exacerbated system response delay and misjudgment probability, forming a technical paradox of increasing data volume and decreasing decision-making timeliness, and with the battery aging problem highlighted in industrial and commercial scenarios. Therefore, how to establish a real-time perception mechanism for the internal electrochemical stability of the battery to avoid the systematic safety risks caused by speculative control has become a technical problem to be solved by the application. SUMMARY

[0005] The application provides an industrial and commercial energy storage battery management system, which mainly aims to solve the problem of control strategy mismatch caused by the lack of real-time perception of the internal state of the battery under high dynamic conditions.

[0006] To achieve the above purpose, the application provides an industrial and commercial energy storage battery management system, which comprises:

[0007] A power management module configured to superimpose a micro-disturbance signal with a frequency of 100 kHz on the main current of the battery charging and discharging by controlling the MOSFET switch on the main loop of the battery charging and discharging;

[0008] A voltage acquisition module configured to acquire the battery terminal voltage response sequence in real time at a sampling frequency of 1 MHz during the superposition of the micro-disturbance signal;

[0009] The state self-identification module is configured to: calculate the approximate entropy value of the voltage response sequence based on the battery terminal voltage response sequence obtained by the voltage acquisition module, where the approximate entropy value characterizes the degree of order of the voltage response sequence and is obtained through the following relationship: ,in, The total length of the voltage sequence. Given similarity tolerance Below, the length is The natural logarithmic mean of the frequency of occurrence of the mode in the voltage response sequence; the current operating mode of the battery is determined by comparing the approximate entropy value with a first entropy threshold and a second entropy threshold. The operating mode includes: stable mode, critical mode or unsteady mode.

[0010] The control decision module is configured to select and execute the corresponding charge and discharge control strategy from a predetermined control strategy library based on the battery operating mode determined by the state self-identification module. The control strategies include: when the operating mode is a stable mode, allowing charge and discharge operations within the nominal range; when the operating mode is a critical mode, performing operations to limit the charge and discharge power; and when the operating mode is a non-steady-state mode, performing protective operations and issuing a warning.

[0011] Preferably, the state self-identification module is configured to: determine the current operating mode of the battery as a stable mode when the approximate entropy value is lower than the first entropy threshold; determine the current operating mode of the battery as a critical mode when the approximate entropy value is greater than or equal to the first entropy threshold and less than the second entropy threshold; and determine the current operating mode of the battery as an unsteady mode when the approximate entropy value is greater than or equal to the second entropy threshold.

[0012] Preferably, the control decision module is configured to perform the operation of limiting charge and discharge power when the current operating mode of the battery is a critical mode, including limiting the charge and discharge power of the battery to 70% of the rated power.

[0013] Preferably, the control decision module is configured to perform protective operations when the current operating mode of the battery is an unsteady mode, including: suspending the high-current charging and discharging operation of the battery and generating a warning signal.

[0014] Preferably, the system further includes: a wake-up trigger module configured to monitor the macroscopic voltage change rate and macroscopic current change rate of the battery at a sampling frequency of ten times per second; when the macroscopic voltage change rate or macroscopic current change rate exceeds a predetermined macroscopic change rate threshold, triggering the voltage acquisition module and the status self-identification module to enter the active working state; when the macroscopic voltage change rate and macroscopic current change rate remain below the macroscopic change rate threshold for five seconds, causing the voltage acquisition module and the status self-identification module to enter the sleep state.

[0015] Preferably, the system further comprises: an input purification gateway module configured to synchronously collect the current signal and the voltage signal of the battery, and extract the total harmonic distortion or peak value of the high-frequency alternating current component from the current signal to generate an electromagnetic pollution index; the input purification gateway module is further configured to dynamically adjust the filtering strength of the voltage response sequence obtained by the voltage acquisition module according to the electromagnetic pollution index, and the filtering strength adjustment includes: using light filtering when the pollution index is low, using medium filtering when the pollution index is moderate, and using strong filtering and temporarily increasing the approximate entropy judgment threshold of the state self-recognition module when the pollution index is heavy.

[0016] Preferably, the system further comprises: a decision arbitration module configured to start a verification window with a duration of two hundred milliseconds when the control decision module issues a high-risk instruction to execute a large power limit or emergency shutdown; the decision arbitration module is further configured to read the temperature value of the thermistor close to the battery tab at a frequency of one thousand times per second during the verification window, and calculate the temperature rise rate; when the temperature rise rate does not exceed a predetermined temperature rise threshold, the high-risk instruction is denied, and the control decision module is instructed to execute a secondary protection strategy, and the state self-recognition module is further configured to: in a low-temperature high-rate working condition, analyze the energy distribution of the battery terminal voltage response sequence in the frequency band of zero point one hertz to ten hertz in parallel; when the energy proportion in the frequency band exceeds a predetermined baseline threshold, the approximate entropy algorithm is suspended, and a zero-crossing point detection method of the second derivative of voltage is used to judge the running mode of the battery.

[0017] Preferably, the system further comprises: a health state detection module configured to apply a zero point one C discharge micro-current pulse lasting one second through the power management module when the battery is in a stable mode and the duration exceeds five minutes; the health state detection module is further configured to collect the voltage recovery curve of the battery for several seconds after the pulse ends, and extract the time taken for the voltage to recover to 63.2 percent of its pre-disturbance value as the relaxation time; the control decision module calculates the battery health state correction factor in real time according to the comparison of the relaxation time and the factory baseline relaxation time of the battery, and corrects all power-related control thresholds according to the correction factor.

[0018] Preferably, the health state correction factor is determined by the following relationship: , wherein, is the baseline relaxation time when the battery is factory-finished, is the currently measured relaxation time.

[0019] A commercial and industrial energy storage battery management method, the method comprising the following steps:

[0020] Step a, superimpose a micro-disturbance signal with a frequency of 100 kHz on the battery charging and discharging main current by controlling the switching device of the power management module; during the superimposition of the micro-disturbance signal, the battery terminal voltage response sequence is collected in real time by the voltage collection module at a sampling frequency of 1 MHz;

[0021] Step b, calculate the approximate entropy value of the voltage response sequence based on the battery terminal voltage response sequence by the state self-identification module, wherein the approximate entropy value represents the ordered degree of the voltage response sequence;

[0022] Step c, determine the current operation mode of the battery according to the comparison result of the approximate entropy value and a first entropy threshold and a second entropy threshold by the state self-identification module, the operation mode including: stable mode, critical mode or non-stable state mode;

[0023] Step d, select and execute the corresponding charging and discharging control strategy from the predetermined control strategy library according to the determined battery operation mode by the control decision module, the control strategy including: when the operation mode is stable mode, allow the execution of the charging and discharging operation within the nominal range; when the operation mode is critical mode, execute the operation of limiting the charging and discharging power; when the operation mode is non-stable state mode, execute the protective operation and issue a warning.

[0024] Compared with the prior art, the beneficial effects of the present application are:

[0025] 1. By analyzing the micro-disturbance excitation and the ordered degree of the voltage response sequence, the system bypasses the dependence of the traditional equivalent model on the historical data, directly captures the instantaneous stability of the internal electrochemical environment of the battery, and this mechanism based on signal micro-texture enables the controller to perceive the transient polarization effect formed by the charging and discharging history in the battery, thereby actively adjusting the power boundary before drastic working condition switching, and avoiding the risk of overcharging and overdischarging caused by model lag.

[0026] 2. The macroscopic change rate monitoring module and the microscopic entropy value calculation module form an asymmetric synergy: when the battery is in a gentle working condition, the system monitors the macroscopic parameter change with extremely low power consumption; only when a drastic charging and discharging demand is detected, the high-frequency sampling and entropy value operation are activated, this resource scheduling mechanism based on working condition dynamics reduces the continuous load of the edge computing device while maintaining the sensitivity of the system, so that the core algorithm can stably run on the affordable hardware; the input purification gateway identifies the intensity of external electromagnetic interference through current harmonic characteristics, and dynamically adjusts the voltage signal processing strategy accordingly, in a strong interference scene, the system automatically enhances the filtering strength and synchronously raises the entropy value determination threshold, so that the state recognition mechanism has environmental adaptability, this design converts electromagnetic noise from the interference source into an environmental variable perceived by the system, which helps to ensure the decision robustness in complex industrial scenes.

[0027] 3. When triggering high-power limit or emergency shutdown instruction, the system carries out secondary verification through millisecond temperature rise rate detection: if the battery tab temperature does not show abnormal rise, the original protection instruction is denied and the secondary strategy is executed, the coupling verification mechanism of the electrochemical entropy change signal and the thermodynamic response avoids unnecessary shutdown caused by single path misjudgment, and improves the system availability; the standardized current pulse is applied in the resting window of the battery stable mode, the internal ion diffusion efficiency is inversely deduced through the relaxation time constant, and the power control threshold is dynamically corrected, the mechanism enables the system to automatically track the boundary condition attenuation caused by battery aging, and helps to keep the performance release of young battery and the safety protection of aged battery at an optimal balance point. BRIEF DESCRIPTION OF DRAWINGS

[0028] Fig. 1 It is a core processing flow diagram of the commercial energy storage battery management system of the application.

[0029] Fig. 2 It is a battery terminal voltage comparison curve diagram of the system and the traditional system in the discharging process.

[0030] Fig. 3 It is a high-risk instruction verification and arbitration flow timing diagram based on temperature rise rate of the application.

[0031] The purpose of the application, the functional characteristics and the advantages will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0032] It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application.

[0033] The commercial energy storage battery management system and method disclosed by the technical solution is based on a core loop composed of a power management module, a voltage acquisition module, a state self-identification module and a control decision module, and cooperates with a wake-up trigger module, an input purification gateway module, a decision arbitration module and a health state detection module to jointly build a multi-level, self-adaptive battery state monitoring and dynamic safety control system. The overall operation logic of the system starts from detecting the internal electrochemical state of the battery by actively applying a small perturbation signal, then quantifies the internal stability based on the entropy value of the voltage response sequence, and finally executes the control strategy accurately matched with the current stability mode, while realizing resource optimization, environment adaptation, decision verification and aging self-correction closed-loop management through peripheral cooperative modules.

[0034] In view of the prior art relying on equivalent circuit model for state estimation, its model updating speed is difficult to match the instantaneous changes of the internal electrochemical state of the battery under high dynamic conditions, thereby causing the control strategy to mismatch the actual carrying capacity of the battery. To directly perceive the immediate stability of the internal electrochemical environment of the battery, the working process of the system starts from the power management module, which is configured to superimpose a micro-disturbance signal with a frequency of on the charge and discharge current of the main loop by controlling the MOSFET switch on the main loop of the battery charge and discharge; at the same time, the voltage acquisition module collects the battery terminal voltage at a high density in real time at a sampling frequency much higher than the disturbance frequency , so as to capture a voltage response sequence that can finely reflect the response of the internal electrochemical environment of the battery to high-frequency excitation; this procedure externalizes the stability of the microscopic dynamic processes such as ion migration and charge transfer in the battery which cannot be directly observed, into a time series containing rich state information and directly available for analysis, thereby bypassing the dependence of traditional equivalent models on historical data; after obtaining the voltage response sequence, to quantify its order degree in a model-free manner and identify the battery operation mode, the core calculation of the system is performed by the state self-identification module, which introduces the approximate entropy algorithm and calculates the approximate entropy value using the following relationship , , where is the total length of the voltage sequence, is the preset pattern matching length, which is calibrated to 2, and the similarity tolerance is dynamically set to 0.15 times the standard deviation of the voltage response sequence by a deterministic procedure to adapt to the voltage fluctuation baseline under different working conditions, is the natural logarithm mean of the frequency of the vector pattern with a length of in the sequence under this tolerance; a lower approximate entropy value represents a highly repetitive and ordered voltage response sequence, indicating that the internal electrochemical environment of the battery is stable, otherwise, a higher approximate entropy value indicates an increase in randomness and complexity of the sequence, indicating that the internal stability is decreasing, this mechanism enables the system to directly understand the immediate health status of the battery.

[0035] Further, the state self-identification module determines the operating state of the battery by comparing the calculated approximate entropy value with a first entropy threshold and a second entropy threshold, thereby determining the operating state of the battery as stable mode, critical mode or unstable mode; the two entropy thresholds are determined by performing full-cycle experiments on the battery in a laboratory controlled environment from normal working condition to overcharge and overdischarge boundary, continuously recording the approximate entropy value distribution of each stage from functional integrity to the eve of thermal runaway, and combining statistical analysis methods to determine the critical point that can best distinguish different stability states; when the approximate entropy value is lower than the first entropy threshold, the battery is determined to be in stable mode; when the approximate entropy value is greater than or equal to the first entropy threshold and less than the second entropy threshold, it is determined to be in critical mode; when the approximate entropy value is greater than or equal to the second entropy threshold, it is determined to be in unstable mode; accordingly, the control decision module calls and executes the corresponding strategy from the predetermined control strategy library according to the mode recognition result: for stable mode, the battery is allowed to perform charging and discharging operation within the nominal power range; for critical mode, the system will actively intervene and forcibly constrain the charging and discharging power to 70% of the rated power; and once in unstable mode, the highest level of protection is triggered, the high-current charging and discharging operation of the battery is suspended, and a warning signal is generated; in order to balance high sensitivity and low power consumption, and to avoid unnecessary calculation power consumption caused by continuous high-frequency sampling and entropy value calculation during stable operation of the battery, the system is configured with a wake-up trigger module; the module monitors the macro voltage and current change rate of the battery at a sampling frequency of ten times per second, and when either change rate exceeds a predetermined macro change rate threshold, for example, when the current change slope exceeds 20% of the rated current of the battery per second, it is judged that the battery will enter high dynamic condition, at which time the voltage sampling module and the state self-identification module are triggered to enter the active working state; otherwise, if the macro voltage change rate and the macro current change rate continuously fall below the macro change rate threshold for five seconds, it is considered that the battery has returned to stable, and the above two modules enter the dormant state; this asymmetric cooperative mechanism maintains the instantaneous response capability to sudden events while reducing the average power consumption of the system.

[0036] Considering the electromagnetic complexity of the industrial and commercial energy storage system deployment environment may contaminate the high-frequency voltage response sequence, the system integrates an input purification gateway module to ensure the robustness of the decision; the gateway synchronously collects the current and voltage signals of the battery, and synthesizes an electromagnetic pollution index by analyzing the total harmonic distortion or peak value of the high-frequency alternating current component in the current signal; the system divides the environment into low, medium, and heavy pollution levels according to this index, and dynamically adjusts the signal processing strategy: in low pollution, a lightweight second-order Butterworth filter is used; in moderate pollution, a medium-strength fourth-order Chebyshev filter is switched; in a heavy pollution environment, a strong sixth-order elliptical filter is used, and on this basis, the first and second entropy thresholds are simultaneously adjusted by 15% in proportion, which changes the environmental noise from a passive interference source to an active sensing system adjustment variable, thereby helping to suppress false positives in a strong noise background.

[0037] To prevent unnecessary protective shutdown caused by a single decision path due to extreme working conditions or sensor abnormalities, the system sets up a decision arbitration module as a safety redundancy mechanism; when the control decision module issues a high-risk instruction to execute a large power limit or emergency shutdown, the arbitration module immediately starts a verification window with a duration of two hundred milliseconds; during this window, the module reads the temperature value of the thermistor close to the battery tab at a frequency of one thousand times per second and calculates its temperature rise rate; only when the temperature rise rate exceeds a preset temperature rise threshold, for example, 0.5 degrees Celsius per second, indicating that the electrochemical instability has indeed converted into a significant thermodynamic response, the original high-risk instruction is confirmed for execution; if the temperature rise rate is not abnormal, the arbitration module rejects the instruction and instructs the control decision module to execute the secondary protection strategy, thereby building a redundant safety confirmation logic based on the cross-validation of electrochemical and thermodynamic signals, improving the overall availability of the system.

[0038] To solve the problem of dynamic changes in performance and safety boundaries of the battery due to aging in the whole life cycle, the system also includes a health state detection module for self-correction of control parameters; the module uses the resting window of the battery in the stable mode and with a duration of more than five minutes, and applies a zero-point-one C discharge micro-current pulse for one second through the power management module; after the pulse ends, the module collects the voltage recovery curve of the battery for several seconds with high precision, and extracts the time taken for the voltage to recover to 63.2% of its pre-disturbance value as the current relaxation time ; the control decision module then calculates a real-time health state correction factor through the relationship , where is the baseline relaxation time of the battery as calibrated at the factory; finally, the system adjusts the correction factor The application is applied to all power-related control thresholds, so that the control strategy can continuously adapt to the actual carrying capacity of the battery throughout its life cycle. In addition, in specific working conditions such as low-temperature high-rate processes, the state self-recognition module also configures parallel analysis logic to deal with the risk of lithium precipitation becoming the main contradiction. In this specific scenario, the module synchronously analyzes the energy distribution of the battery terminal voltage response sequence in the frequency band of zero point one hertz to ten hertz. When the energy proportion in this frequency band exceeds a predetermined baseline threshold, the system determines that the risk of lithium precipitation has increased dramatically. At this time, the approximate entropy algorithm is suspended, and the zero-crossing point detection method of the second derivative of the voltage, which is more sensitive to lithium precipitation, is used to determine the running mode of the battery. This algorithm switching mechanism for specific risk scenarios further improves the accuracy and coverage of system state recognition.

[0039] Embodiment 1: In a container-type energy storage power station providing grid frequency modulation and peak-valley arbitrage services for a large industrial park, the power station needs to provide power support for the instantaneous start-stop of a large rolling mill in the park. The challenge of this working condition is that the rolling mill generates a large amount of steep discharge impact when starting, and during the intermittent period of operation, the energy storage system needs to use the low-valley electricity price of the grid for rapid charging. When the energy storage system has just completed a high-rate charging and the battery state of charge is close to saturation, the park control system issues an instruction requiring the energy storage power station to immediately provide peak power discharge for the rolling mill about to start. At this moment, the traditional management system based on macro parameters can only determine that the power is sufficient, but cannot perceive the electrode polarization effect formed in the battery during the high-rate charging process, which has not relaxed, thereby hiding the risk of electrode voltage sudden drop and triggering protection shutdown during discharge. In this working condition, the wake-up trigger module of the system monitors that the macro current change rate has exceeded the predetermined macro change rate threshold, and immediately activates the high-frequency sampling and state self-recognition module, switching the system from low-power standby state to full-time perception state. Due to the moderate level of electromagnetic pollution index of the grid environment caused by the operation of frequency conversion equipment in the industrial park, the input purification gateway module has automatically switched to a moderate intensity fourth-order Chebyshev filtering strategy before entropy calculation, thereby supplying a high-fidelity voltage response sequence to the state self-recognition module. The state self-recognition module receives the sequence and calculates its approximate entropy value, which is found to be higher than the first entropy threshold but lower than the second entropy threshold. The system determines that the battery has entered a critical mode, and its decision basis has changed from the battery state of charge estimation to the direct quantification of its internal electrochemical stability.

[0040] Accordingly, the control decision module does not execute the full-power discharge instruction issued by the main control system according to the identification result of the critical mode, but limits the maximum discharge power of the battery to 70% of the rated power before the rolling mill load is actually connected. The power limitation instruction triggers the decision arbitration module, which does not exceed the predetermined temperature rise threshold in the two-hundred-millisecond verification window period. The stable thermodynamic signal rejects the emergency stop instruction, so that the power limitation, a forward-looking intervention measure, is executed. The rolling mill is finally started at the limited power, and the terminal voltage of the energy storage battery appears a short-term drop, but does not reach the low-voltage protection threshold, thereby avoiding an unplanned shutdown due to the mismatch between the control strategy and the actual carrying capacity of the battery. After the power-limited discharge process ends, the internal polarization of the battery is buffered, and the subsequent approximate entropy calculation result returns below the first entropy threshold. The battery enters a stable mode, and the control decision module removes the power limitation and restores the nominal charge and discharge capacity of the battery. When the energy storage system runs stably for more than a preset time, the health state detection module automatically triggers a health state detection process using the resting window. By applying a small current pulse of zero point one C and measuring the relaxation time , the system calculates the current health state correction factor , and uses the factor to lower all power-related control thresholds. This architecture places real-time sensing of internal electrochemical stability at the core of the control logic and builds a collaborative mechanism around it for resource adaptation, environment adaptation, decision self-validation, and health self-correction, forming a closed-loop system with a control boundary that can be reconstructed in real time according to the internal state of the battery.

[0041] Example 2: To objectively verify the ability of the present technical solution to identify the potential unstable state of the battery under high dynamic conditions, especially the early warning lead of the present technical solution relative to the traditional voltage threshold monitoring method, a controlled test is constructed in this example; the test platform is composed of a programmable high-precision battery tester, a temperature-controlled test box, a commercial ternary lithium-ion battery monomer, a standard battery management system as a control group, and a battery management system prototype equipped with all the functional modules of the present invention; among them, the standard battery management system only relies on the battery terminal voltage and state of charge for protective judgment, while the prototype of the present invention runs the approximate entropy calculation and multi-module collaborative decision mechanism in parallel; in the test design, the frequency of the perturbation signal is a key setting parameter, its value directly affects the detection sensitivity of the specific electrochemical process and the signal-to-noise ratio of the hardware implementation; the core factor affecting the setting of this parameter is the electrochemical impedance spectrum characteristics of the battery at different frequencies, and the characteristic time constant of different electrochemical processes; the essence of the technical trade-off is to balance between ensuring that the signal can detect the electrode interface state closely related to the stability of the battery and avoiding the serious attenuation of high-frequency signals and the dramatic increase in the complexity of the measurement circuit; given that the electrochemical process response related to the stability of the electrode interface is mostly concentrated in the frequency band of several kilohertz to several hundred kilohertz, in order to ensure high sensitivity to the changes in the interface state while avoiding the noise of low-frequency conditions and the hardware implementation bottleneck of ultra-high frequency, the frequency of the perturbation signal in this test is set to , which is a non-limiting exemplary value for the ternary lithium-ion battery system in the frequency band.

[0042] The test procedure is designed to simulate the condition that the battery undergoes high-rate charging and then immediately undergoes high-current discharging; first, the battery monomer at 25 degrees Celsius is placed and adjusted to a nominal state of charge of 50%; then, the battery is charged to a state of charge of 95% by the battery tester at a rate of 2C; after the end of the charging phase, without a resting period, a 3C rate constant current discharge instruction is immediately applied, and the key parameters and decision outputs of the two management systems are continuously recorded; during the test, the running track and decision sequence of the two systems are recorded to be different; at the end of the charging phase, the voltage and state of charge readings of the control group system are always within the normal working interval, and until the discharge instruction is issued, the battery state is still good and allows 3C rate discharging; in contrast, the approximate entropy value calculated by the state self-identification module of the prototype of the present invention shows a continuous rise in the second half of the charging phase, and before the charging is completed and the discharge instruction is issued, the entropy value has triggered the system to switch from the stable mode to the critical mode.

[0043] Table 1: Comparison of key node data for the test.

[0044]

[0045] Referring to Table 1, the variation trend of the test data reflects the inherent working mechanism of the system. The increase of the approximate entropy value corresponds to the dramatic increase of the lithium ion concentration gradient in the electrode material and the electrolyte, the intensified electrode surface polarization, and the transformation of the battery internal electrochemical environment from a relatively uniform and orderly state to an unbalanced and randomly disturbed disordered state. The deterioration of the internal stability is captured by the increase of the complexity of the high-frequency voltage response sequence through the approximate entropy algorithm. The test results show that the battery state recognition method based on the perturbation response approximate entropy value can identify the electrochemical unstable state caused by the intensified internal polarization before the abnormality of the battery macroscopic electrical parameters, and accordingly execute the protective control prior to the traditional voltage threshold method.

[0046] Example 3: This example is in conjunction with Figs. 1 to 3 a commercial energy storage battery management system and method, such as Fig. 1As shown, first, in the input signal part, the system receives signal input from the energy storage battery unit, while introducing a thermistor temperature rise rate monitoring signal for subsequent temperature rise verification link, the core processing module is composed of multiple functional sub-modules, first, the power management module MOSFET controls the superposition of 100kHz micro-disturbance signal for applying 100kHz micro-disturbance signal on the battery charging and discharging main loop; the voltage acquisition module 1MHz high frequency sampling voltage response sequence acquisition module acquires voltage response sequence at 1MHz frequency, which is used for subsequent entropy value calculation; the wake-up trigger module macroscopic change rate monitoring resource self-adaptive control module monitors the macroscopic voltage and current change rate of the battery at low frequency sampling, and activates the high-frequency sampling logic only when severe changes occur; the input purification gateway module electromagnetic pollution index calculation dynamic filtering adjustment module extracts the electromagnetic pollution index according to the real-time current signal, and dynamically adjusts the filtering strategy of the voltage signal; the state self-recognition module approximate entropy value calculation running mode determination module calculates the approximate entropy value of the collected voltage response sequence, and determines the battery running mode according to the preset threshold; the decision arbitration module temperature rise rate verification cross-validation mechanism module verifies whether the temperature rise rate of the thermistor exceeds the threshold in real time when the control strategy triggers a high-risk instruction to determine whether the instruction is executed; the health state detection module relaxation time measurement correction factor calculation module applies a micro-current pulse and measures the relaxation time in the stable mode, thereby calculating the battery health state correction factor; the control decision module mode matching strategy power limitation / protection control module adjusts the control strategy according to the current running mode, combined with the health factor, and executes the corresponding power limitation or protection measures, in addition, for the voltage second derivative zero crossing detection situation under low temperature and high rate working condition, the system parallelly enables voltage second derivative zero crossing point analysis to enhance the state recognition accuracy under the risk working condition of lithium extraction, finally, in the control output part, according to the different running modes, output the corresponding control commands: if in stable mode, execute nominal power charging and discharging; if in critical mode, execute power limitation to 70%; if in non-steady state mode, execute protection shutdown + warning.

[0047] As shown in Fig. 2 From the figure, it can be seen that between 0s and 1.5s, the battery terminal voltage curves of the system of the application and the traditional system are relatively close, represented by solid line with circle and dashed line with square respectively, then the voltage of the traditional system rapidly drops after 1.5s, and decreases to near 2.8V at 2s and tends to be stable, showing that it fails to identify the polarization effect caused by charging in advance, resulting in a sudden voltage drop at the moment of discharging, while the system of the application relies on approximate entropy value calculation, mode determination and power limitation mechanism, and can still maintain the battery terminal voltage slowly decreasing and stably maintaining above 3.4V after 2s, which is significantly better than the traditional system, avoiding the risk of voltage sudden drop triggering protection shutdown.

[0048] As shown in Fig. 3As shown, first, when the control decision module determines to be in a stable mode according to the state mode, the nominal power charging and discharging is allowed; if it is determined to be a critical mode, a control strategy of limiting the power to 70% of the rated value is issued; if it is a non-steady state mode, the protection shutdown is directly triggered and a warning signal is sent to the early warning system, at this time, the control decision module will issue a high-risk instruction, trigger the decision arbitration module to start a 200ms verification window, and read the tab temperature in the window period by the decision arbitration module in the form of sampling once per millisecond, the temperature value is collected and returned by the temperature sensor, then the temperature rise rate is calculated, if the temperature rise rate exceeds the predetermined value, it is confirmed to execute the high-risk instruction, and the power management module executes the pause of large current charging and discharging; if the temperature rise rate is normal, the high-risk instruction is vetoed, and the control decision module is instructed to execute the secondary protection strategy.

[0049] In order to ensure the decision reliability of the system in the whole life cycle and complex working conditions, the internal key parameters and specific risk identification logic are determined through a series of offline calibration processes and online adaptive procedures. As for the predetermined temperature rise threshold value relied on by the decision arbitration module, it is not a fixed value, but in the battery monomer design and finalization stage, a controllable small internal short circuit is applied to the sample battery, or a thin film heater close to the battery is used to simulate a quantitative heat generation rate, so as to establish a transfer function relationship from the internal abnormal heat generation power to the temperature rise rate of the thermistor at the tab. The temperature rise rate value caused by the heat generation power which is confirmed as the irreversible thermal runaway starting point by the battery thermal model is determined as the predetermined temperature rise threshold value. For the electromagnetic pollution index used by the input purification gateway module to quantify the electromagnetic environment interference, the calculation procedure is designed as a deterministic process of normalizing and weighting the total harmonic distortion and the peak-to-peak value of the high-frequency alternating component. The system first measures the current total harmonic distortion and the peak-to-peak value of the high-frequency alternating component in the reference clean working condition in an electromagnetic shielding environment and peak-to-peak value , then in actual operation, the real-time collected and are combined into a dimensionless electromagnetic pollution index through the relationship , wherein the weight factor and According to the sensitivity difference of different types of noise to the approximate entropy algorithm of this type of battery, the noise is calibrated through offline injection noise experiment, so that the index can reflect the actual interference level to the core algorithm.

[0050] And in the low temperature high rate condition, in order to deal with the risk of lithium precipitation, the voltage second derivative zero crossing point detection method is triggered according to the energy ratio baseline threshold value, which is obtained by charging the sample battery at different rates in the low temperature test box, and the lithium metal counter electrode connected in series with the battery is monitored in situ. When the deposition platform of lithium is first detected on the counter electrode potential, the energy ratio of the battery terminal voltage response sequence in the frequency band of zero point one hertz to ten hertz at this moment is recorded, and this value is taken as the predetermined baseline threshold value for triggering the algorithm switching. Once the algorithm is triggered, the state self recognition module counts the number of sign changes of the voltage response sequence filtered by the band pass filter in a fixed time window, and the count value is the number of zero crossing points. A count value close to zero represents a smooth voltage curve, corresponding to a stable charging process, and is defined as a stable mode. With the occurrence of lithium precipitation, the uneven charge distribution on the electrode surface will cause slight oscillation of the voltage, resulting in an increase in the number of zero crossing points. When the count value exceeds a value significantly related to lithium precipitation risk calibrated in the above in situ monitoring experiment, the system determines the current operating mode as critical mode or unstable state mode, and transfers the control decision module to perform the corresponding power limitation or charging cutoff operation. A second derivative calculation based on central difference method is performed, that is, Wherein is the sampling point index; the module counts the number of sign changes of sequence in a fixed time window, and the count value is the number of zero crossing points. A count value close to zero represents a smooth voltage curve, corresponding to a stable charging process, and is defined as a stable mode. With the occurrence of lithium precipitation, the uneven charge distribution on the electrode surface will cause slight oscillation of the voltage, resulting in an increase in the number of zero crossing points. When the count value exceeds a value significantly related to lithium precipitation risk calibrated in the above in situ monitoring experiment, the system determines the current operating mode as critical mode or unstable state mode, and transfers the control decision module to perform the corresponding power limitation or charging cutoff operation.

[0051] In the first deployment and integration of a new battery energy storage unit in the commercial battery management system, the system needs to perform a pre baseline calibration procedure. In this procedure, the new battery unit needs to be adjusted to a stable state with a state of charge in the range of 50% to 60%, a difference between internal temperature and ambient temperature less than 2 degrees Celsius and a standing time of more than one hour. Then, a baseline relaxation time automatic measurement program is triggered through the maintenance interface of the system. The program repeatedly performs a zero point one C discharge small current pulse and calculates its relaxation time until the standard deviation of the relaxation time values obtained by the latest ten measurements is less than a preset convergence threshold value. When the program terminates, the statistical average value of the ten measurements is taken as the baseline relaxation time of the specific battery unit and is written into the non volatile storage area of the system. Based on the baseline relaxation time obtained in situ , the control decision module in the system takes this value as the health state correction factor for all subsequent calculations The calculated benchmark is used to initialize and calibrate all power-related control thresholds in the predetermined control strategy library. This is a standardization engineering debugging process performed before the system is formally put into operation, which tracks the aging state of the battery management system and anchors it on the initial health state benchmark of each energy storage unit, which is calibrated on site. The commercial and industrial energy storage battery management system checks the integrity of the key sensors at each power-on start. The system reads the resistance value of the thermistor relied on by the decision arbitration module. If the reading is outside the upper and lower limits of the normal working resistance range defined in the specification of the sensor, the system determines that the sensor is in an open or short circuit fault state. At this time, the system will lock the veto function of the decision arbitration module, allowing all high-risk instructions to be directly released in subsequent operation, and simultaneously generating a highest level alarm that needs to be reset by manual. After completing the sensor verification, the procedure enters the parameter self-tuning stage of the wake-up trigger module. The system retrieves and analyzes the recorded macroscopic current time series in the last complete running cycle, calculates the 95th percentile value of the current change rate in the sequence . Then, the system automatically sets the macroscopic change rate threshold of the wake-up trigger module to 1.2 times the statistical value, and sets the duration threshold of the high-frequency sampling and entropy value operation in the sleep state by the relationship , where is the preset maximum sleep time, is a fixed coefficient calibrated according to the system power consumption and response speed requirements.

[0052] It is obvious to those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0053] Finally, it should be noted that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A commercial energy storage battery management system, characterized by, The system comprises: a power management module configured to superimpose a micro-perturbation signal on the main current of the battery charging and discharging by controlling the MOSFET switch on the main charging and discharging circuit of the battery; a voltage acquisition module configured to acquire a voltage response sequence of the battery in real time during the superimposition of the micro-perturbation signal; The state self-identification module is configured to: based on the battery terminal voltage response sequence obtained by the voltage acquisition module, calculate an approximate entropy value of the voltage response sequence, wherein the approximate entropy value represents an ordered degree of the voltage response sequence, and is obtained through the following relationship: wherein, is a total length of the voltage sequence, is a natural logarithm average value of a frequency of a pattern with a length of in the voltage response sequence under a given similarity tolerance ; and determine a current operating mode of the battery according to a comparison result of the approximate entropy value with a first entropy threshold value and a second entropy threshold value, wherein the operating mode includes: a stable mode, a critical mode or a non-steady state mode. a control decision module configured to select and execute a corresponding charging and discharging control strategy from a predetermined control strategy library according to the operating mode of the battery determined by the state self-identification module, the control strategy comprising: when the operating mode is a stable mode, allowing the execution of charging and discharging operation within a nominal range; when the operating mode is a critical mode, executing a limited charging and discharging power operation; and when the operating mode is a non-steady state mode, executing a protective operation and issuing a warning.

2. A commercial energy storage battery management system according to claim 1, wherein, The state self-identification module is configured to: when the approximate entropy value is lower than a first entropy threshold, determine that the current operating mode of the battery is a stable mode; when the approximate entropy value is greater than or equal to the first entropy threshold and less than a second entropy threshold, determine that the current operating mode of the battery is a critical mode; and when the approximate entropy value is greater than or equal to the second entropy threshold, determine that the current operating mode of the battery is a non-steady state mode.

3. A commercial energy storage battery management system according to claim 2, wherein, The control decision module is configured to, when the current operating mode of the battery is a critical mode, execute a limited charging and discharging power operation, which includes limiting the charging and discharging power of the battery to 70% of the rated power.

4. A commercial energy storage battery management system according to claim 2, wherein, The control decision module is configured to, when the current operating mode of the battery is a non-steady state mode, execute a protective operation, which includes suspending the high-current charging and discharging operation of the battery and generating a warning signal.

5. A commercial energy storage battery management system according to claim 1, wherein, The system further comprises a wake-up trigger module configured to monitor the macro-voltage change rate and the macro-current change rate of the battery at a sampling frequency of ten times per second; when the macro-voltage change rate or the macro-current change rate exceeds a predetermined macro-change rate threshold, triggering the voltage acquisition module and the state self-identification module to enter an active working state; and after the macro-voltage change rate and the macro-current change rate continuously fall below the macro-change rate threshold for five seconds, causing the voltage acquisition module and the state self-identification module to enter a dormant state.

6. A commercial energy storage battery management system according to claim 1, wherein, The system further comprises an input purification gateway module configured to synchronously acquire the current signal and the voltage signal of the battery, and extract the total harmonic distortion or peak value of the high-frequency alternating current component from the current signal to generate an electromagnetic pollution index; the input purification gateway module is further configured to dynamically adjust the filtering strength of the voltage response sequence obtained by the voltage acquisition module according to the electromagnetic pollution index, the filtering strength adjustment including: using light filtering when the pollution index is low, using medium-intensity filtering when the pollution index is moderate, and using strong filtering and temporarily increasing the approximate entropy determination threshold of the state self-identification module when the pollution index is severe.

7. A commercial energy storage battery management system according to claim 1, wherein, The system further comprises a decision arbitration module configured to initiate a verification window with a duration of 200 milliseconds when the control decision module issues a high-risk instruction to execute a large power limit or an emergency shutdown; the decision arbitration module is further configured to read the temperature value of the thermistor close to the battery tab at a frequency of 1000 times per second during the verification window and calculate the temperature rise rate; when the temperature rise rate does not exceed a predetermined temperature rise threshold, the high-risk instruction is rejected, and the control decision module is instructed to execute a secondary protection strategy; the state self-identification module is further configured to: in a low-temperature high-rate working condition, analyze the energy distribution of the battery terminal voltage response sequence in the frequency band of 0.1 Hz to 10 Hz in parallel; when the energy proportion in the frequency band exceeds a predetermined baseline threshold, the approximate entropy algorithm is suspended, and the zero-crossing point detection method of the second derivative of the voltage is used to determine the running mode of the battery.

8. A commercial energy storage battery management system according to claim 1, wherein, The system further comprises a health state detection module configured to apply a 0.1C discharge micro-current pulse lasting one second through the power management module when the battery is in a stable mode and the duration exceeds five minutes; the health state detection module is further configured to collect the voltage recovery curve of the battery for several seconds after the pulse ends, and extract the time taken for the voltage to recover to 63.2% of its pre-disturbance value as the relaxation time; the control decision module calculates the battery health state correction factor in real time according to the comparison of the relaxation time and the factory baseline relaxation time of the battery, and corrects all power-related control thresholds based on this.

9. A commercial energy storage battery management system according to claim 8, wherein, Health status correction factor is determined by the relationship: wherein, is the baseline relaxation time at the factory of the battery, is the currently measured relaxation time.

10. A method of managing industrial and commercial energy storage batteries, comprising: The method is performed based on the system of claim 1, comprising the following steps: Step a, by controlling the switching device of the power management module, superimpose a micro-disturbance signal with a frequency of 100 kHz on the battery charging and discharging main current; during the superimposition of the micro-disturbance signal, the voltage acquisition module acquires the battery terminal voltage response sequence in real time at a sampling frequency of 1 MHz; Step b, by the state self-identification module, calculate the approximate entropy value of the voltage response sequence based on the battery terminal voltage response sequence, wherein the approximate entropy value represents the ordered degree of the voltage response sequence; Step c, by the state self-identification module, determine the current running mode of the battery according to the comparison result of the approximate entropy value with a first entropy threshold and a second entropy threshold, the running mode including: stable mode, critical mode or non-steady state mode; Step d, by the control decision module, select and execute the corresponding charging and discharging control strategy from the predetermined control strategy library according to the determined running mode of the battery, the control strategy including: when the running mode is stable, allow the execution of charging and discharging operation within the nominal range; when the running mode is critical, execute the operation of limiting the charging and discharging power; when the running mode is non-steady state, execute the protective operation and issue a warning.

Images