Battery protection method, device and equipment and storage medium

Abstract

The invention provides a battery protection method and device, equipment and a storage medium. The method comprises the following steps: acquiring battery state data of each single battery in a battery pack; carrying out difference comparison and threshold cross judgment among the single batteries according to the battery state data, and generating a multi-stage abnormal signal and an abnormal intensity score value; based on the multi-level abnormal signal and the abnormal strength score value, retrieving a corresponding protection level entry from a preset protection strategy library, and generating a multi-level protection instruction; and performing combined control on a current output state corresponding to the battery pack and a circuit breaker according to the multi-level protection instruction. Through the implementation of the scheme of the invention, the state data of each single battery is collected, and difference comparison and multi-level cross judgment are executed among the state data, so that not only can grade division and intensity scoring be carried out on anomalies, but also corresponding response levels can be retrieved from the preset strategy library, and finally a secondary protection scheme based on fault properties and risk levels is formed. And the safety of battery protection can be effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of battery protection, and in particular to a battery protection method, device, equipment and storage medium. BACKGROUND

[0002] In current application fields such as new energy vehicles, electric tools and renewable energy storage systems, the safety and working stability of lithium batteries as the main energy carrier have become key factors for system operation. Taking new energy electric vehicles as a specific application scenario, such systems often rely on the battery management system (BMS) for real-time monitoring and primary protection, but in some sudden failure scenarios (such as power device short circuit, sampling abnormal false alarm, main control logic lag), the traditional primary protection may fail, thereby causing irreversible damage to the battery monomer or the whole system.

[0003] Therefore, in order to improve the reliability of the whole vehicle operation, the existing technology usually uses a hardware circuit breaker for physical isolation when the primary protection mechanism fails to respond in time. However, the existing secondary protection method mostly relies on single-point threshold triggering, lacks deep perception of the internal state difference and fault evolution trend of the battery pack, and is prone to over-protection due to misjudgment. SUMMARY

[0004] The present application provides a battery protection method, device, equipment and storage medium, which is used to solve the problem of over-protection caused by misjudgment when the related technology relies on single-point threshold triggering for secondary protection of the battery.

[0005] The first aspect of the present application provides a battery protection method, which comprises: obtaining battery state data of each monomer battery in a battery pack; performing difference comparison and threshold crossing judgment between the monomer batteries according to the battery state data, generating multi-level abnormal signal and abnormal intensity score value; based on the multi-level abnormal signal and the abnormal intensity score value, retrieving the corresponding protection level entry from a preset protection strategy library to generate multi-level protection instructions; jointly controlling the current output state and the circuit breaker of the battery pack according to the multi-level protection instructions.

[0006] Optionally, in the first implementation manner of the first aspect of the present application, the step of obtaining the battery state data of each monomer battery in the battery pack comprises: synchronously sampling the electrical data of each monomer battery in the battery pack to generate a first data set; According to the environmental temperature data in the sampling period and the real-time temperature value of the single battery, a temperature compensation coefficient is determined, and the first data set is corrected through the temperature compensation coefficient to generate a second data set; According to the voltage change trend value of the single battery in the preset historical period, a difference function between the current voltage sampling value is determined; According to the difference function, the voltage parameter of the second data set is corrected again to generate a third data set; The third data set is structurally bound to the corresponding single battery number and the sampling time stamp to generate the battery state data of each single battery.

[0007] Optionally, in the second implementation manner of the first aspect of the present application, the step of generating a multi-level abnormal signal and an abnormal intensity score value by comparing and judging the threshold value according to the battery state data between the single batteries, comprises: The electrical data of the single battery is compared with the corresponding preset critical threshold value, and a first level abnormal signal is generated according to the corresponding comparison result; According to the battery state data, the parameter change rate of the single battery in the continuous sampling period is obtained, the risk abnormality of the parameter change rate is identified, and a second level abnormal signal is generated; The state data of the single battery in the current sampling period is compared with the historical stable running data, and it is identified whether the single battery has behavior deviation, and a third level abnormal signal is generated when the deviation degree reaches the preset deviation threshold; The abnormal intensity score value corresponding to the first level abnormal signal, the second level abnormal signal and the third level abnormal signal is determined through abnormal level weighting logic.

[0008] Optionally, in the third implementation manner of the first aspect of the present application, the step of retrieving the corresponding protection level item from the preset protection strategy library based on the multi-level abnormal signal and the abnormal intensity score value, and generating a multi-level protection instruction, comprises: According to the level identifier and the abnormal intensity score value corresponding to the multi-level abnormal signal, the protection strategy type to be called is determined through index mapping logic, and a primary response instruction item corresponding to the protection strategy type is generated; By comparing the action conditions in the primary response instruction item with the battery state data, a strategy item group whose matching degree meets the preset determination standard is identified, and a multi-level response candidate set is generated; The abnormal information in the multi-level abnormal signal is matched with the response mode of the multi-level response candidate set to generate a multi-level protection instruction covering the current fault situation.

[0009] Optionally, in a fourth implementation form of the first aspect of the application, the step of jointly controlling the current output state of the battery pack and the circuit breaker according to the multi-level protection instruction comprises: generating a first control instruction to adjust the current output state of the corresponding battery pack according to the action level field and the power limiting ratio of the multi-level protection instruction; determining whether a trigger condition for performing a circuit breaker state switching is met according to the fault severity level and the duration threshold of the multi-level protection instruction, generating a disconnection command and controlling the circuit breaker of the electrical connection path to disconnect when the trigger condition is met; acquiring actual response data of the circuit breaker switching behavior by sampling the battery output voltage and the bus current before and after the circuit breaker switching; comparing the response data with the target state of the multi-level protection instruction to determine whether the circuit breaker disconnection conforms to a preset logical path; if the preset logical path is met, generating a closing command by starting a timing mechanism and a state determination mechanism, and controlling the circuit breaker to switch from the disconnected state to the closed state according to the closing command.

[0010] Optionally, in a fifth implementation form of the first aspect of the application, the method further comprises, after the step of jointly controlling the current output state of the battery pack and the circuit breaker according to the multi-level protection instruction: determining a sensing control mode of the target state and generating a corresponding sampling control instruction according to the sensing strategy parameter and the action execution record associated in the multi-level protection instruction; dynamically adjusting a non-critical channel sampling period in the battery pack by scheduling a channel priority parameter of the sampling configuration instruction; applying a periodic pulse signal to the sensing circuit by controlling the duty cycle and the output mode of the current output to obtain a periodic battery state according to the adjusted sampling period; determining a voltage noise interval corresponding to the current battery working phase according to the voltage stability data of the periodic battery state; generating a periodic configuration table for activating the sensing circuit by performing fluctuation boundary analysis on the voltage noise interval.

[0011] Optionally, in a sixth implementation form of the first aspect of the application, the method further comprises: determining a standby running interval according to the system wake-up frequency and the power output state in an activation period corresponding to the periodic configuration table; generating a fluctuation reference range by counting the battery voltage fluctuation amplitude and the sampling channel trigger density of the standby running interval. According to the fluctuation reference range and the sensing channel grouping strategy of the sampling control instruction, a trigger weight mapping table of the night sampling channel is constructed, and a trigger threshold of a corresponding channel is adjusted according to the trigger weight mapping table; When no abnormal signal exceeding the trigger threshold is detected within a preset sampling period, the trigger path configuration of the sensing circuit is updated according to the trigger threshold.

[0012] The second aspect of the application provides a battery protection device, the battery protection device is used for realizing the battery protection method, and the battery protection device comprises: An acquisition module is configured to acquire battery state data of each single battery in a battery pack. A judgment module is configured to perform difference comparison and threshold crossing judgment among the single batteries according to the battery state data, to generate multi-level abnormal signals and abnormal intensity score values. A generation module is configured to retrieve corresponding protection level entries from a preset protection strategy library based on the multi-level abnormal signals and the abnormal intensity score values, to generate multi-level protection instructions. A control module is configured to jointly control a current output state and a circuit breaker of the battery pack according to the multi-level protection instructions.

[0013] The third aspect of the embodiments of the application provides an electronic device, comprising a memory and a processor, wherein the processor is used to execute a computer program stored in the memory, and when the processor executes the computer program, each step of the battery protection method provided in the first aspect of the embodiments of the application is implemented.

[0014] The fourth aspect of the embodiments of the application provides a computer readable storage medium, which stores a computer program, and when the computer program is executed by a processor, each step of the battery protection method provided in the first aspect of the embodiments of the application is implemented.

[0015] In summary, according to the battery protection method, device, equipment and storage medium provided in the application, the battery state data of each single battery in the battery pack is acquired; difference comparison and threshold crossing judgment are performed among the single batteries according to the battery state data, and a multi-level abnormal signal and an abnormal intensity score value are generated; based on the multi-level abnormal signal and the abnormal intensity score value, a corresponding protection level item is retrieved from a preset protection strategy library, and a multi-level protection instruction is generated; and the current output state and the circuit breaker corresponding to the battery pack are jointly controlled according to the multi-level protection instruction. Through the implementation of the application, by collecting the state data of each single battery and performing difference comparison and multi-level crossing judgment thereon, not only can the abnormality be graded and scored, but also the corresponding response level can be retrieved from the preset strategy library, and finally a secondary protection scheme based on the fault nature and risk level is formed, which can effectively provide the safety of battery protection. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flowchart of a battery protection method provided by an embodiment of the application is shown. Figure 2 A program module diagram of a battery protection device provided by an embodiment of the application is shown. Figure 3 A structure diagram of an electronic device provided by an embodiment of the application is shown. DETAILED DESCRIPTION

[0017] In order to make the application purposes, features and advantages of the application more obvious and easy to understand, the technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0018] In order to solve the problem that the related art depends on single-point threshold triggering to perform secondary protection on the battery, which is easy to misjudge and cause over-protection, an embodiment of the application provides a battery protection method, as shown in Figure 1 A flowchart of a battery protection method provided by the embodiment is shown, and the battery protection method includes the following steps: Step 110, acquiring battery state data of each single battery in a battery pack.

[0019] In the embodiment, the parameters such as voltage, current and temperature are synchronously sampled by the multi-channel sampling assembly, and the sampling values are temperature compensated and trend corrected in combination with the current environmental temperature and historical data, so that the single battery state data closer to the real state is obtained. All sampling results are provided with the cell number and time label for subsequent calling. The process not only covers the collection of conventional static electrical parameters, but also embeds dynamic change identification logic, which can retain the original data and corrected data in parallel structure at the same time of sampling, and ensure that each single battery state has comparability and traceability.

[0020] In an optional implementation of the embodiment, the step of obtaining the battery state data of each single battery in the battery pack includes: synchronously sampling the electrical data of each single battery in the battery pack to generate a first data set; determining a temperature compensation coefficient according to the environmental temperature data and the real-time temperature value of the single battery in a sampling period, and correcting the first data set by the temperature compensation coefficient to generate a second data set; determining a difference function between the current voltage sampling value and the voltage variation trend value of the single battery in a preset historical period; secondarily correcting the voltage parameter of the second data set according to the difference function to generate a third data set; and structurally binding the third data set with the corresponding single battery number and sampling time stamp to generate the battery state data of each single battery.

[0021] Specifically, in the present embodiment, synchronous sampling refers to simultaneous collection of physical quantities such as voltage, current, temperature, etc. of all objects to be measured at the same time point, avoiding data mismatch problems caused by inconsistent sampling timing. In specific applications, a multi-channel analog-to-digital converter (ADC) or a high-speed polling controller can be used to drive all channels in parallel to collect data, thereby constructing a complete time-consistent data frame. Taking the power battery of an electric vehicle as an example, when multiple series-connected cells undergo dynamic changes under high-rate discharge conditions, if there is a delay in the parameter collection of different cells, it will lead to system errors and false imbalance states. Therefore, synchronous sampling can ensure that the first data set obtained has time domain consistency, so that the data between different cells is comparable. After the electrical parameter sampling is completed, the raw data needs to be corrected according to the real-time temperature condition. The temperature compensation coefficient is used to reflect the response characteristics of the cell voltage and current under different temperature conditions, and its calculation is based on the relationship between the ambient temperature and the cell body temperature at the time of sampling. In actual engineering, the table lookup method or mathematical fitting method can be used to obtain the deviation relationship curve of temperature and voltage, and the temperature correction value of each sampling point is calculated based on this. Taking a lithium iron phosphate cell as an example, it has the characteristic of voltage platform moving down under low temperature conditions. If the raw voltage value is directly used for abnormal judgment, it may be misjudged as over-discharge or capacity attenuation. Therefore, by correcting the first data set with the temperature compensation coefficient, a second data set can be generated, and each voltage and current value in the new data set has taken into account the temperature factor, making it closer to the performance under standard working conditions and ensuring data consistency under different environments. In order to further correct the influence of long-term evolution factors on the current data, a trend difference function needs to be introduced to compare the current voltage value with the change trend in the historical period. The trend value is a characteristic parameter obtained by extracting the slope and regression fitting of the voltage change data of the same cell in the past stable period, used to describe the long-term evolution behavior of the cell. When there is a significant difference between the current sampling value and the trend value, it means that the cell may be in a non-steady state fluctuation or have an abnormal resistance growth. In the high-frequency start-stop scene of electric tools, some cells may show inconsistent slow voltage collapse after frequent discharge, and if the trend deviation evaluation is not introduced, the traditional judgment method will ignore the potential risk. Therefore, by constructing the difference function between the current voltage value and the trend value, the degree of deviation between the individual state and the historical model can be quantified, providing pre-support for abnormal identification. After obtaining the difference function, the voltage parameters in the second data set are corrected again to further eliminate the data deviation caused by cell aging, cycle difference, etc. This correction adds the value of the difference function as a dynamic compensation to the voltage value that has been temperature corrected, obtaining a third data set that integrates the current state and historical characteristics.Taking the cascade utilization battery pack in the energy storage power station as an example, due to different sources and different histories, each cell may exhibit different response characteristics even in the same environment. Simple temperature correction cannot distinguish such long-term differences. Only by superimposing trend difference analysis can the voltage performance be more accurately analyzed, so that the final voltage data achieves uniformity and comparability in both time and state dimensions.

[0022] After the above correction processing is completed, each set of data in the third data set is structurally bound with the unique number of the single battery to which it belongs and the corresponding sampling time stamp to form a structured data unit. This binding operation realizes the tracking and backtracking management of data by encapsulating data values, device identifiers, and time information into a unified data structure. In the intelligent battery management system, this structured expression not only helps to track the running evolution trajectory of the cell, but also can be connected with the external communication module to realize event recording and remote diagnosis. At the application level, this binding structure significantly improves the usability and integration efficiency of data, so that the sampling data not only has decision value in the current period, but also provides high-dimensional adjustable data basis for future data reconstruction, health assessment, and service life prediction.

[0023] Step 120, difference comparison and threshold cross judgment between single batteries according to battery state data, generating multi-level abnormal signal and abnormal intensity score value.

[0024] In this embodiment, the state data of each single battery is compared with other cell data item by item, and cross-validated with the preset threshold or historical statistical extreme value to complete the abnormal identification. The comparison rules include but are not limited to threshold determination logic, trend slope analysis function, and feature deviation judgment rule based on historical samples. In the abnormal identification process, each determination will generate a first to third level abnormal signal, and the abnormal signal quantity, position, intensity and persistence are combined and weighted to finally obtain the corresponding abnormal intensity score.

[0025] In an optional implementation of the embodiment, the step of generating multi-level abnormal signal and abnormal intensity score value according to the difference comparison and threshold cross judgment between single batteries according to battery state data, includes: comparing the electrical data of the single battery with the corresponding preset critical threshold, and generating a first level abnormal signal according to the corresponding comparison result; obtaining the parameter change rate of the single battery in the continuous sampling period according to the battery state data, identifying the risk abnormality of the parameter change rate, and generating a second level abnormal signal; comparing the state data of the single battery in the current sampling period with the historical stable running data, identifying whether the single battery has behavior deviation, and generating a third level abnormal signal when the deviation degree reaches the preset deviation threshold; determining the abnormal intensity score value corresponding to the first level abnormal signal, the second level abnormal signal and the third level abnormal signal through abnormal level weighting logic.

[0026] Specifically, in the present embodiment, the operating state of a single battery can be expressed by its electrical data such as voltage, current and temperature. Comparing these parameters with the corresponding preset critical threshold is a direct method to identify whether the battery is in an out-of-limit state. The critical threshold is a fixed or dynamic boundary value set according to the battery specifications, material properties and typical working conditions, and usually includes safety limits such as maximum temperature, minimum voltage, maximum charge and discharge current, etc. For example, the voltage threshold of a ternary lithium battery can be set between 2.8V and 4.2V, and any instantaneous measurement above or below this interval is considered to have safety risks. In the comparison process, each parameter is independently judged by a threshold comparison logic circuit or software function. When any threshold is exceeded, a first-level abnormal signal is generated and marked as a critical out-of-limit type of abnormality. Unlike methods that rely on average values, using individual threshold comparison can capture the risk of single-point extreme values of individual batteries, avoiding the masking of local abnormalities by overall trends. For gradual faults that cannot be immediately exposed by static comparison, parameter change rate needs to be introduced for dynamic analysis. The change rate is the change slope obtained by dividing the parameter difference in the continuous sampling period by the time interval, which reflects the transient evolution trend of the battery state. For example, although the temperature has not exceeded the limit at a certain moment, if the temperature rise rate exceeds the predetermined value, it will indicate a potential tendency of thermal runaway. Similarly, during discharging, if the voltage drop rate abnormally accelerates, it may mean that the battery capacity is unbalanced or the internal resistance is rapidly rising. The identification of such rate characteristics can be achieved by a first-order difference algorithm or a sliding window slope regression, and the abnormality recognition process generates a second-level abnormal signal when the change rate exceeds the threshold. Since the change rate analysis is forward-looking, it can provide early warning before the parameter reaches the absolute out-of-limit, so this method has higher sensitivity and prediction ability than static threshold judgment under dynamic load conditions. The identification of behavior deviation does not rely on fixed thresholds, but compares the current battery state with its historical running trajectory through pattern recognition methods. Historical stable data refers to a sample set composed of feature values of the battery in a long-term stable running state, including static parameter mean value, change trend and correlation structure. Difference comparison can be achieved by constructing a state vector and using functions such as Euclidean distance, Manhattan distance or cosine similarity to evaluate the deviation between the current state and the reference state. For example, in a gradient battery application, some batteries are running in a certain voltage and temperature range for a long time. If the current state deviates significantly, it should be judged as abnormal even if it is not out of limit. By comparing the deviation with the preset behavior deviation threshold, a third-level abnormal signal is output once the set boundary is exceeded. This method has individual adaptability and pattern recognition ability compared with traditional judgment, and can accurately identify non-structural faults caused by aging or process differences. After the generation of the three types of abnormal signals, the weighted model can be used to integrate the influence degree to form an abnormal intensity score value.The score value is not simply added, but is weighted according to factors such as abnormal level, duration, cell location, etc. For example, if a cell located in the thermal center has a second-level abnormality, its score value may be higher than that of a first-level out-of-limit signal located in the edge area. The weighting function can adopt the form of exponential decay, interval mapping or fuzzy logic to normalize and fuse different categories and intensities of abnormalities, and output an abnormality intensity score value with risk interpretation.

[0027] Step 130, based on the multi-level abnormality signal and the abnormality intensity score value, retrieving the corresponding protection level entry from the preset protection strategy library to generate multi-level protection instructions.

[0028] In this embodiment, the system analyzes the acquired abnormality signal level and score result, and determines the type of protection strategy to be called according to the preset mapping table. The strategy library contains response entries under different fault levels, each entry containing response actions, current limiting strategies, power-off logic and recovery criteria. The retrieval process of the protection strategy is based on multiple fields such as abnormal position, cell number, score value and duration, and performs condition matching and priority selection to select the instruction set covering the current state from multiple matching strategies. Once the strategy entry is selected, the corresponding action logic is generated according to the response path, and all action logics are packaged into control command structures, including action trigger conditions, execution methods, expected state values and other fields, which are used for subsequent response process calling.

[0029] In an alternative embodiment of the present embodiment, the step of retrieving the corresponding protection level entry from the preset protection strategy library based on the multi-level abnormality signal and the abnormality intensity score value to generate multi-level protection instructions includes: determining the type of protection strategy to be called and generating the primary response instruction entry corresponding to the protection strategy type according to the level identifier corresponding to the multi-level abnormality signal and the abnormality intensity score value through index mapping logic; identifying the strategy entry group whose matching degree meets the preset determination standard by comparing the action conditions in the primary response instruction entry with the battery state data, and generating a multi-level response candidate set; and performing rule matching between the abnormal information in the multi-level abnormality signal and the response mode of the multi-level response candidate set to generate multi-level protection instructions covering the current fault situation.

[0030] Specifically, in the present embodiment, after the multi-level abnormal signal and the corresponding abnormal intensity score value are generated, the information can be mapped to the preset protection strategy type through index mapping logic. The mapping process determines the index path according to multiple dimensions such as abnormal level identification, score value numerical interval, and cell number belonging to. The mapping logic can be constructed by a multi-dimensional hash table, a mapping matrix, or a state transition diagram to quickly locate the response strategy type under different input combinations. For example, when the system receives a third-level structural abnormal signal from the core hot area, and the score value exceeds the critical threshold, the mapping logic will automatically point to the strategy type with the label of "critical thermal slope + center position priority disconnection", and generate the corresponding primary response instruction entry. The entry is a response template classified by protection level in the strategy library, including strategy call entry, corresponding action content, response threshold, priority label, and state identification field. The entry structure defines the response mode that should be triggered in this type of fault scenario, ensuring consistency between the subsequent response logic and the identification result. After the primary response instruction entry is selected, its content needs to be compared with the real-time state data of the current battery pack field by field to determine the adaptability. The comparison content mainly includes current output level, cell number, abnormal duration, voltage stability, and other condition fields. Taking the discharge abnormality as an example, if the strategy entry requires a pressure drop greater than 500 millivolts within 10 seconds and the abnormality is concentrated in the middle area, and the current sampling data only meets part of the conditions, then the strategy will not be included in the candidate set; on the contrary, if the trigger conditions of another strategy entry completely cover the current state characteristics, it is determined as a high matching degree entry. The comparison mechanism can be realized by Boolean operation logic, fuzzy logic matching, or weight mapping function to support condition tolerance comparison outside the given threshold, thereby forming a multi-level response candidate set. Each candidate entry has an operation path for a specific scenario, including current limiting, group disconnection, transient monitoring, thermal maintenance, and other instruction combinations. After filtering the strategy entries in the candidate set, they need to be further matched with the specific fault structure characteristics contained in the abnormal signal. The rule matching process is based on the structure information of the abnormal signal, including the signal source channel, fault type label (such as temperature jump, pressure difference anomaly, internal resistance deviation, etc.), and the combination structure of the abnormal signal. The matching rule takes the response mode field as the core reference, combines the execution priority and the action impact range, and selects the strategy combination entry that best fits the current fault situation in the structure dimension. Taking the slope imbalance of high-voltage cells as an example, if the abnormal signal comes from multiple physically adjacent channels and the abnormal type is consistent, the matching rule will preferentially select the strategy instruction with regional linkage mechanism, rather than the single-point protection type, thereby generating a set of multi-level protection instructions with coherent action paths, close logical relationship, and complete coverage. Each protection instruction includes structured fields such as action level identification, execution trigger condition, recovery criterion, controlled object list, and execution time limit, so that the final output is not only an action call, but also a complete response process embedded with logical constraints.

[0031] Step 140, jointly controlling the current output state of the battery pack and the circuit breaker according to the multi-level protection instruction.

[0032] In this embodiment, the power supply path and the circuit breaker state are jointly controlled according to the formed action instruction structure. The control logic sends current limiting, voltage reducing or closing instructions to the power output link, and sends opening or closing control signals to the circuit breaker path, to complete the linkage execution of the protection response action. At a certain fault level, only part of the path will enter the soft switching state, and in higher level responses, physical isolation will be triggered. The entire control process tracks the battery pack output end voltage, current and circuit breaker state feedback signals in real time, and performs action closed loop comparison through state sampling to verify whether the response meets the logical expectation. All feedback data will also be used for subsequent state judgment and possible self-recovery logic evaluation, thereby forming a closed protection response loop.

[0033] In an optional embodiment of the present embodiment, the step of jointly controlling the current output state of the battery pack and the circuit breaker according to the multi-level protection instruction includes: generating a first control instruction to adjust the current output state of the corresponding battery pack according to the action level field and the power limiting ratio value of the multi-level protection instruction; judging whether the trigger condition for switching the state of the circuit breaker is met according to the fault severity level and the duration threshold of the multi-level protection instruction, and generating a disconnection command and controlling the circuit breaker of the electrical connection path to disconnect when the trigger condition is met; obtaining actual response data of the circuit breaker switching behavior by sampling the battery output voltage and bus current before and after the circuit breaker switching; comparing the response data with the target state of the multi-level protection instruction to judge whether the circuit breaker disconnection meets the preset logical path; if it meets the preset logical path, generating a closing instruction by starting a timing mechanism and a state determination mechanism, and controlling the circuit breaker to switch from the disconnected state to the closed state according to the closing instruction.

[0034] Specifically, in this embodiment, after the multi-level protection instruction is generated, the system needs to extract the control quantity according to the action level field and the power limit ratio in it to construct the first control instruction to adjust the current output state of the battery pack. The action level field is usually represented by multi-level identifiers such as L1, L2, L3, etc., corresponding to different risk levels and response intensities, while the power limit ratio refers to the proportion of the battery's allowed output power to the maximum output power under the current protection state, expressed in percentage or unitized value, for example, 0.6 means that the power output is limited to 60% of the maximum power. After receiving such parameters, the control system determines the target current value through table lookup or direct calculation, and issues control instructions to the current adjustment module in the battery pack output path with this value. This control process is usually completed by DC-DC converters, PWM controllers or intelligent bus drive modules to achieve the effect of limiting current intervention in the fault propagation area while maintaining the power supply capability. Taking a new energy public transportation system as an example, if the front axle motor drive battery detects a moderate abnormality, limiting the power to 70% of the rated value can prevent the system from cascading due to heat load expansion, while avoiding the operation interruption caused by direct power-off, compared with the traditional full-off strategy, this mechanism shows higher system availability control capability. While executing the current limiting control, the controller also needs to read the fields about the fault severity level and duration threshold in the multi-level protection instruction, and determine whether the trigger condition of the circuit breaker switching is met through a logical judgment process. The severity level field is used to represent the range and urgency of abnormal influence, and the duration threshold field is used to define when the forced physical isolation operation needs to be performed. When a certain abnormal signal level is L3 and it has lasted more than 10 seconds in the sampling period, the system considers that the fault has entered an unrecoverable stage, generates a disconnect command through the trigger logic block when the condition is met, and calls the circuit breaker control interface on the electrical connection path to execute the disconnect action. When the circuit breaker performs switching, it needs to select the optimal time point in combination with the power path load state to avoid short-term impact and arc occurrence, so before the disconnect command is executed, it also needs to evaluate the slope according to the real-time current value to complete the mechanical switching at the steady state point. This process is completed by the drive chip built-in the circuit breaker. Taking the main power battery of an electric forklift as an example, when a certain cell heat runaway develops rapidly, if the system only limits current without timely cutting off the power supply path, it is easy to cause heat diffusion, while the hardware disconnect action can build a truly physical isolation path, which is different from logical current limiting, and has the ability to completely interrupt energy flow, forming an independent fault plane guarantee. After the circuit breaker completes the switching action, to verify the execution effect and action accuracy of the control instruction, the system needs to sample the battery output voltage and bus current before and after the circuit breaker action, and the sampling signals are generated by the voltage collection module and the current transformer synchronously and stored in the cache area with time stamp. The actual response data reflects the physical state difference on both sides of the circuit before and after the circuit breaker is disconnected, which is the only basis for evaluating whether the circuit breaker action achieves the expected target.The response data includes key indicators such as whether the bus current before disconnection is effectively cleared, whether the output voltage presents a step drop, etc. Subsequently, the control system compares the response data with the expected state field in the multi-level protection instruction, such as the expected bus current less than 0.1A, voltage drop not less than 90%, etc. When the response data meets all instruction targets, it is confirmed that the circuit breaker action has been completed according to the logical path, otherwise an abnormal action alarm will be issued and the re-evaluation link will be entered. Unlike the existing method of only using disconnection signal feedback as the basis for execution, this structural comparison mechanism provides a closed-loop verification logic based on physical quantity feedback, significantly enhancing the certainty of system protection action. After the circuit breaker disconnection action meets the target state condition, the control logic will enter the recovery evaluation stage. In this stage, the controller starts the timing mechanism and state determination mechanism according to the post-disconnection cell temperature, voltage fluctuation degree, current zero value retention time, etc. to build a judgment model according to the logical back table. When the timer records that the static stable retention period has exceeded the minimum recovery time threshold, and all sampled signals are back to the safe interval, the system generates a closing instruction and sends a reclosing control signal to the circuit breaker control interface. The circuit breaker switches from the open state to the closed state, restoring the electrical path connection. The recovery action and the initial disconnection action have independent control logic and are controlled by redundant safety conditions to avoid secondary faults caused by false closing.

[0035] In an optional implementation of the embodiment, after the step of jointly controlling the current output state of the battery pack and the circuit breaker according to the multi-level protection instruction, the method further includes: determining a sensing control mode of the target state and generating a corresponding sampling control instruction according to the sensing strategy parameters and the action execution record associated in the multi-level protection instruction; dynamically adjusting the non-critical channel sampling period in the battery pack by scheduling the channel priority parameters of the sampling configuration instruction; applying a periodic pulse signal to the sensing circuit by controlling the duty cycle and output mode of the current output according to the adjusted sampling period to obtain a periodic battery state; determining the voltage noise interval corresponding to the current battery working stage according to the voltage stability data of the periodic battery state; and generating a periodic configuration table for activating the sensing circuit by performing fluctuation boundary analysis on the voltage noise interval.

[0036] Specifically, in the present embodiment, the multi-level protection instruction contains two types of information, sensing strategy parameters and action execution records. The former indicates the preset state of the sensor working mode under different fault conditions, such as sampling accuracy, cycle frequency and signal type, etc. The latter reflects the specific execution behavior in the previous strategy response process, such as current limiting effective time, breaker action number or fault duration length, etc. By jointly analyzing the sensing strategy parameters and action execution records, the control logic can construct the state classification of the current system, and match the target sensing control mode accordingly, such as switching from high-frequency sampling mode to energy-optimized sampling mode or maintaining in fault observation sensitive mode. The controller generates a sampling control instruction according to the structure definition of the sensing control mode, which includes fields such as sampling channel index, sampling window length, target accuracy level, activation condition, etc. Through the combination of these parameters, the sampling scheduler and data acquisition circuit are driven into an adaptive state. Taking the grid-side energy storage device as an example, during the battery pack stability detection after interruption recovery, the sensing mode can automatically enter the low-noise gravity sampling state. Compared with the fixed frequency working mode, this adaptive sampling instruction mechanism improves the responsiveness of sampling allocation. After the generation of the sampling control instruction, the channel priority parameter needs to be scheduled to dynamically adjust the sampling period of non-critical channels in the battery pack. The channel priority parameter indicates the data importance and sampling urgency of each sampling channel in the current working state. Critical channels such as hot center area, abnormal signal source point, critical capacity cell, etc. need to maintain high sampling frequency, while non-critical channels such as edge stable area, historical non-abnormal trajectory monomer, etc. can reduce sampling density to release sampling resources and reduce redundant energy consumption. By dynamically adjusting the sampling period, different timer frequencies or sampling trigger thresholds can be allocated to each channel to realize an asynchronous scheduling mechanism. This mechanism can be completed by a multi-level sampling frequency divider or programmable timing logic to configure the independent sampling clock of the sampling channel. In the light-load energy storage system operation, when the main abnormality focuses on the middle cluster of cells, the sampling period of peripheral monomers can be adjusted from 100ms to 500ms, thereby releasing more sampling windows for core area use, enhancing the abnormal coverage capability of the system in the key time window, and reducing the overall system resource conflict. After the sampling period is adjusted, the system will apply a periodic pulse signal to the sensing circuit by controlling the duty cycle and output mode of the current source according to the updated sampling frequency parameter. The duty cycle refers to the proportion of time that the current output is in the active state within a period, and the output mode determines whether the current is in continuous mode or pulse mode. In low-power sampling design, pulse power supply can significantly reduce standby power consumption, while avoiding heat accumulation or zero drift of sampling devices caused by long-time current flow. The amplitude and width of the pulse signal are dynamically set by the controller according to the sampling type (such as voltage collection, current monitoring or internal resistance measurement), and are applied to the excitation input end of the sensing circuit to trigger the corresponding sampling module to collect periodic battery state data.As the periodic battery status data is continuously collected, the system can statistically analyze the voltage stability data therein to determine the voltage noise interval corresponding to the current battery operating phase. The voltage stability data is composed of small fluctuations in the sampled signal sequence, and the range of its changes reflects the natural deviation characteristics of the battery under the current load, temperature, and internal resistance conditions. This noise interval is extracted by methods such as range analysis, variance calculation, or wavelet transform. In an industrial environment, the voltage fluctuation range under normal operation is usually narrow, while irregular peaks or mutations occur under abnormal conditions. By modeling the stability data of different battery cells in multiple time windows, the natural noise bandwidth can be accurately depicted, providing a basis for the subsequent determination of the sampling threshold. Based on the identified voltage noise interval, the system can further analyze its fluctuation boundaries and generate a periodic configuration table for activating the sensing circuit in combination with the target sensing requirements. The configuration table contains fields such as the trigger period, delay period, and warning response interval for each type of sampling behavior, guiding the sensing system's work rhythm in a specific noise environment. Fluctuation boundary analysis can use adaptive gating logic, i.e., dynamically setting the trigger criterion, so that the sensing circuit enters a high-frequency sampling state when the voltage fluctuation exceeds the statistical tolerance, and maintains a low-frequency detection when the fluctuation is within the noise band. For example, in a large communication base station power system, using the configuration table can ensure that the sensing circuit is only activated in the key window, thereby reducing system redundancy, allowing the system to be functional while maintaining low energy consumption. Compared with the traditional equal-period sampling method, this approach has better resource allocation accuracy and dynamic sensing capability.

[0037] In an alternative embodiment of the present embodiment, the standby operation interval is determined according to the system wake-up frequency and power output state within the activation period corresponding to the periodic configuration table; the fluctuation reference range is generated by statistically analyzing the battery voltage fluctuation amplitude and sampling channel trigger density in the standby operation interval; the trigger weight mapping table of the night sampling channel is constructed according to the fluctuation reference range and the sensing channel grouping strategy of the sampling control instruction, and the trigger threshold of the corresponding channel is adjusted according to the trigger weight mapping table; when no abnormal signal exceeding the trigger threshold is detected within the preset sampling period, the trigger path configuration of the sensing circuit is updated according to the trigger threshold.

[0038] Specifically, in the embodiment, the sensing circuit activation period defined in the period configuration table contains multiple trigger time points for sensing the battery state. By obtaining the wake-up frequency information of the system in each activation period, i.e., the number of times the system switches from a low-power state to a running state, and jointly judging the current power output state, it can be inferred whether the system is in a smooth running state outside the significant load change in the period. If the power output is maintained in a low load or constant load state for a long time in the activation period, and the system wake-up frequency is below the low threshold, it can be determined that the system enters the standby running interval. The standby running interval is a state of non-work or long static standby, characterized by low energy consumption, slow electrical response, and few disturbance signals. After identifying the standby running interval, by continuously sampling the voltage values of each single battery in the battery pack in this time period, the voltage fluctuation amplitude range is extracted, and the number of trigger events generated by the sensing channel is counted, the trigger density distribution of different channels in the low disturbance environment can be obtained. The voltage fluctuation amplitude is used to measure the fluctuation boundary of the stability of the standby interval battery, and the channel trigger density reflects the active degree of the channel response to subtle fluctuations, and the combination of the two can establish an electrical fluctuation reference range representing the sensitivity of the channel. According to the above fluctuation reference range, the sensing channels contained in the sampling control instruction are grouped according to sensitivity, priority or sampling frequency strategy, and different groups of channels are assigned weight coefficients to construct a trigger weight mapping table. The trigger weight mapping table is a data structure based on channel grouping, in which the trigger weight of each channel represents the reasonable threshold intensity interval of being triggered in the night or standby stage. By using the mapping table, the trigger threshold in the original sampling control logic can be reset, so that the high-sensitivity channel obtains a more relaxed response tolerance band in the static stage, and the low-priority channel is given a higher suppression threshold, thereby adjusting the trigger behavior of the sensing system in the non-working state as a whole. With the advancement of the sampling period, when the system does not detect any channel with abnormal signals exceeding the adjusted threshold in a plurality of consecutive activation periods, the trigger path configuration of the sensing circuit can be updated based on the existing trigger threshold parameter set. The trigger path configuration refers to the decision diagram of which sensing channels in the control system are activated in which state. By updating the configuration table, part of the channels can enter a low-frequency trigger or suspended trigger state at night, reducing the number of circuit switchings and sampling power consumption, while retaining the response capability of the key path.

[0039] According to the battery protection method provided in the scheme, battery state data of each single battery in the battery pack is acquired; difference comparison and threshold cross judgment are performed between the single batteries according to the battery state data, multi-level abnormal signal and abnormal intensity score value are generated; corresponding protection level entries are retrieved from a preset protection strategy library based on the multi-level abnormal signal and the abnormal intensity score value, multi-level protection instructions are generated; and the current output state and the circuit breaker corresponding to the battery pack are jointly controlled according to the multi-level protection instructions. Through the implementation of the scheme, by collecting the state data of each single battery and performing difference comparison and multi-level cross judgment therebetween, not only can the abnormality be graded and intensity scored, but also the corresponding response level can be retrieved from the preset strategy library, and finally a secondary protection scheme based on the fault property and the risk level is formed, which can effectively provide the safety of battery protection.

[0040] Figure 2 A battery protection device is provided for the embodiments of the present application. The battery protection device can be used to implement the battery protection method in the foregoing embodiments. As shown in the figure, the battery protection device mainly includes: Figure 2 An acquisition module 10 is configured to acquire battery state data of each single battery in a battery pack. A judgment module 20 is configured to perform difference comparison and threshold cross judgment between the single batteries according to the battery state data, and generate multi-level abnormal signal and abnormal intensity score value. A generation module 30 is configured to retrieve corresponding protection level entries from a preset protection strategy library based on the multi-level abnormal signal and the abnormal intensity score value, and generate multi-level protection instructions. A control module 40 is configured to jointly control the current output state and the circuit breaker corresponding to the battery pack according to the multi-level protection instructions.

[0041] In an optional implementation of the present embodiment, the acquisition module is specifically configured to: synchronously sample electrical data of each single battery in the battery pack to generate a first data set; determine a temperature compensation coefficient according to environmental temperature data in a sampling period and a real-time temperature value of the single battery, and correct the first data set by the temperature compensation coefficient to generate a second data set; determine a difference function between a current voltage sampling value and a voltage variation trend value of the single battery in a preset historical period; secondarily correct a voltage parameter of the second data set according to the difference function to generate a third data set; and structureally bind the third data set, a corresponding single battery number and a sampling time stamp to generate the battery state data of each single battery.

[0042] ​In an optional implementation of the embodiment, the determining module is specifically configured to: compare the electrical data of the single battery with a corresponding preset critical threshold, and generate a first-level abnormal signal according to a corresponding comparison result; obtain a parameter change rate of the single battery in a continuous sampling period according to the battery state data, identify a risk abnormality of the parameter change rate, and generate a second-level abnormal signal; compare the state data of the single battery in the current sampling period with historical stable operation data, identify whether the single battery has behavior deviation, and generate a third-level abnormal signal when a deviation degree reaches a preset deviation threshold; and determine an abnormal intensity score value corresponding to the first-level abnormal signal, the second-level abnormal signal and the third-level abnormal signal through abnormal level weighting logic.

[0043] In an optional implementation of the embodiment, the generating module is specifically configured to: determine a protection strategy type to be called through index mapping logic according to the level identifier and the abnormal intensity score value corresponding to the multi-level abnormal signal, and generate a primary response instruction item corresponding to the protection strategy type; identify a strategy item group with a matching degree satisfying a preset determination standard through condition comparison between action conditions in the primary response instruction item and the battery state data, and generate a multi-level response candidate set; and perform rule matching between abnormal information in the multi-level abnormal signal and response modes of the multi-level response candidate set, and generate a multi-level protection instruction covering a current fault situation.

[0044] In an optional implementation of the embodiment, the control module is specifically configured to: generate a first control instruction to adjust a current output state of the corresponding battery pack according to an action level field and a power limiting ratio value of the multi-level protection instruction; determine whether a trigger condition for executing a circuit breaker state switching is satisfied according to a fault severity level and a duration threshold of the multi-level protection instruction, generate an opening command when the trigger condition is satisfied, and control a circuit breaker of an electrical connection path to be opened through the opening command; obtain actual response data of the circuit breaker switching behavior by sampling battery output voltage and bus current before and after the circuit breaker switching; compare the response data with a target state of the multi-level protection instruction, and determine whether the circuit breaker opening conforms to a preset logical path; if the preset logical path is conformed to, generate a closing instruction through a starting timing mechanism and a state determination mechanism, and control the circuit breaker to be switched from an open state to a closed state according to the closing instruction.

[0045] In an optional implementation of the embodiment, the control module is further configured to: determine a sensing control mode of the target state and generate a corresponding sampling control instruction according to the sensing strategy parameter associated in the multi-level protection instruction and the action execution record; dynamically adjust a non-critical channel sampling period in the battery pack by scheduling a channel priority parameter of the sampling configuration instruction; apply a periodic pulse signal to the sensing circuit by controlling a duty cycle and an output mode of the current output according to the adjusted sampling period to obtain a periodic battery state; determine a voltage noise interval corresponding to a current battery working stage according to voltage stability data of the periodic battery state; and generate a periodic configuration table of the activated sensing circuit by performing fluctuation boundary analysis on the voltage noise interval.

[0046] In an optional implementation of the embodiment, the control module is further configured to: determine a standby running interval according to a system wake-up frequency and a power output state in an activation period corresponding to the periodic configuration table; generate a fluctuation reference range by statistically analyzing a battery voltage fluctuation amplitude and a sampling channel trigger density of the standby running interval; construct a trigger weight mapping table of the night sampling channel according to the fluctuation reference range and a sensing channel grouping strategy of the sampling control instruction, and adjust a trigger threshold of a corresponding channel according to the trigger weight mapping table; and update a trigger path configuration of the sensing circuit according to the trigger threshold when no abnormal signal exceeding the trigger threshold is detected within a preset sampling period.

[0047] According to the battery protection device provided in the application, battery state data of each single battery in the battery pack is acquired; multi-level abnormal signals and abnormal intensity score values are generated by performing difference comparison and threshold crossing judgment between the single batteries according to the battery state data; a protection level item corresponding to the multi-level abnormal signals and the abnormal intensity score values is retrieved from a preset protection strategy library to generate a multi-level protection instruction; and the current output state and the circuit breaker corresponding to the battery pack are jointly controlled according to the multi-level protection instruction. Through the implementation of the application, the state data of each single battery is collected, and difference comparison and multi-level crossing judgment are performed thereon, which can not only grade and score the abnormality, but also retrieve the corresponding response level from the preset strategy library, and finally form a secondary protection scheme based on the fault property and the risk level, which can effectively provide the safety of battery protection.

[0048] According to the battery protection device provided in the application, Figure 3 An electronic device is provided for the embodiment of the application. The electronic device can be used to implement the battery protection method in the foregoing embodiments, and mainly includes: The system includes a memory 301, a processor 302, and a computer program 303 stored on the memory 301 and executable on the processor 302. The memory 301 and the processor 302 are connected via communication. When the processor 302 executes the computer program 303, it implements the battery protection method described in the foregoing embodiments. The number of processors can be one or more.

[0049] The memory 301 can be a high-speed random access memory (RAM) or a non-volatile memory, such as a disk storage device. The memory 301 is used to store executable program code, and the processor 302 is coupled to the memory 301.

[0050] Furthermore, embodiments of this application also provide a computer-readable storage medium, which may be disposed in the electronic device described in the above embodiments, and the computer-readable storage medium may be as described above. Figure 3 The memory in the illustrated embodiment.

[0051] The computer-readable storage medium stores a computer program that, when executed by a processor, implements the battery protection method described in the foregoing embodiments. Furthermore, the computer-readable storage medium can also be a USB flash drive, a portable hard drive, a read-only memory (ROM), RAM, a magnetic disk, or an optical disk, or any other medium capable of storing program code.

[0052] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0053] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0054] The above-described embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A battery protection method, characterized in that, include: Obtain the battery status data of each individual cell in the battery pack; Based on the battery status data, differences are compared and threshold cross-judgments are made between the individual cells to generate multi-level abnormal signals and abnormal intensity scores. Based on the multi-level abnormal signals and the abnormal intensity score, the corresponding protection level entries are retrieved from the preset protection strategy library to generate multi-level protection instructions; The current output status of the battery pack and the circuit breaker are jointly controlled according to the multi-level protection instructions.

2. The battery protection method according to claim 1, characterized in that, The step of obtaining the battery state data of each individual cell in the battery pack includes: The electrical data of each individual cell in the battery pack are sampled synchronously to generate a first data set; The temperature compensation coefficient is determined based on the ambient temperature data and the real-time temperature value of the individual battery cells within the sampling period, and the first data set is corrected using the temperature compensation coefficient to generate the second data set. Based on the voltage change trend value of the individual battery cell within a preset historical period, determine the difference function between the current voltage sampling value and the current voltage sampling value. The voltage parameters of the second data set are corrected a second time based on the difference function to generate the third data set; The third data set is structurally bound to the corresponding individual battery number and sampling timestamp to generate battery status data for each individual battery.

3. The battery protection method according to claim 2, characterized in that, The step of comparing differences and making threshold cross-judgments among the individual cells based on the battery state data to generate multi-level abnormal signals and abnormal intensity scores includes: The electrical data of the individual battery cell is compared with the corresponding preset critical threshold, and a first-level abnormal signal is generated based on the comparison result. Based on the battery status data, the parameter change rate of the individual battery cell within a continuous sampling period is obtained, the risk anomaly of the parameter change rate is identified, and a second-level anomaly signal is generated. The status data of the individual battery in the current sampling period is compared with the historical stable operation data to identify whether the individual battery has a behavioral deviation. When the deviation reaches the preset deviation threshold, a third-level abnormal signal is generated. The anomaly intensity score corresponding to the first-level anomaly signal, the second-level anomaly signal, and the third-level anomaly signal is determined by anomaly level weighting logic.

4. The battery protection method according to claim 1, characterized in that, The step of retrieving corresponding protection level entries from a preset protection strategy library based on the multi-level abnormal signals and the abnormal intensity scores, and generating multi-level protection instructions, includes: Based on the level identifier and abnormal intensity score corresponding to the multi-level abnormal signal, the type of protection strategy to be invoked is determined through index mapping logic, and a primary response instruction entry corresponding to the protection strategy type is generated. By comparing the action conditions in the primary response instruction entries with the battery status data, a strategy entry group that meets the preset judgment criteria is identified, and a multi-level response candidate set is generated. The abnormal information in the multi-level abnormal signals is matched with the response mode of the multi-level response candidate set according to rules to generate a multi-level protection instruction that covers the current fault situation.

5. The battery protection method according to claim 1, characterized in that, The step of jointly controlling the current output state of the battery pack and the circuit breaker according to the multi-level protection instructions includes: Based on the action level field and power limit ratio of the multi-level protection command, a first control command is generated to adjust the current output state of the corresponding battery pack. Based on the fault severity level and duration threshold of the multi-level protection command, determine whether the triggering conditions for circuit breaker state switching are met. When the triggering condition is met, a disconnect command is generated and the circuit breaker of the electrical connection path is controlled to disconnect through the disconnect command; By sampling the battery output voltage and bus current before and after the circuit breaker switching, the actual response data of the circuit breaker switching behavior is obtained. The response data is compared with the target state of the multi-level protection command to determine whether the circuit breaker disconnection conforms to the preset logic path. If the preset logical path is met, a closing instruction is generated by starting a timing mechanism and a state determination mechanism, and the circuit breaker is controlled to switch from the open state to the closed state according to the closing instruction.

6. The battery protection method according to claim 1, characterized in that, After the step of jointly controlling the current output state of the battery pack and the circuit breaker according to the multi-level protection instructions, the method further includes: Based on the sensing strategy parameters and action execution records associated with the multi-level protection instructions, the sensing control mode of the target state is determined and the corresponding sampling control instructions are generated. The sampling period of non-critical channels in the battery pack is dynamically adjusted by scheduling the channel priority parameters of the sampling configuration command. Based on the adjusted sampling period, by controlling the duty cycle and output mode of the current output, a periodic pulse signal is applied to the sensing circuit to obtain the periodic battery state; Based on the voltage stability data of the periodic battery state, determine the voltage noise range corresponding to the current battery operating stage; By performing fluctuation boundary analysis on the voltage noise range, a period configuration table for activating the sensing circuit is generated.

7. The battery protection method according to claim 1, characterized in that, The method further includes: The standby operation interval is determined based on the system wake-up frequency and power output status within the activation period corresponding to the cycle configuration table. A fluctuation reference range is generated by statistically analyzing the battery voltage fluctuation amplitude and sampling channel trigger density during the standby operation period. Based on the fluctuation reference range and the sensing channel grouping strategy of the sampling control command, a trigger weight mapping table for the night sampling channel is constructed, and the trigger threshold of the corresponding channel is adjusted according to the trigger weight mapping table. When no abnormal signal exceeding the trigger threshold is detected within the preset sampling period, the trigger path configuration of the sensing circuit is updated according to the trigger threshold.

8. A battery protection device, characterized in that, The battery protection device is used to implement the battery protection method according to claim 1, and the battery protection device includes: The acquisition module is used to acquire the battery status data of each individual cell in the battery pack. The judgment module is used to compare differences and cross-judge thresholds between individual cells based on the battery status data, and generate multi-level abnormal signals and abnormal intensity scores. The generation module is used to retrieve the corresponding protection level entries from the preset protection strategy library based on the multi-level abnormal signals and the abnormal intensity score, and generate multi-level protection instructions. The control module is used to jointly control the current output status of the battery pack and the circuit breaker according to the multi-level protection instructions.

9. An electronic device, characterized in that, Includes memory and processor, of which: The processor is used to execute computer programs stored in the memory; When the processor executes the computer program, it implements the steps of the battery protection method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the battery protection method according to any one of claims 1 to 7.

Images