A protection control method for overcharging of a generator lithium battery

Abstract

This invention discloses a protection and control method for overcharging of a generator lithium battery, belonging to the field of lithium battery safety control technology. The method constructs a joint monitoring dataset by acquiring real-time status data from the generator and lithium battery sides. Based on this dataset, it identifies overcharge trigger candidate events and determines the battery's near-state. An overcharge risk assessment result is obtained through fusion calculation of these two data points. Protection levels are then matched based on the risk assessment results. Closed-loop analysis and parameter adjustment are performed on the matching results to form a protection parameter set. Finally, graded protection control is executed to obtain the overcharge protection control result. This invention facilitates proactive overcharge risk control, reduces the probability of lithium battery overvoltage damage, improves the reliability of protection control in unattended scenarios with weak grids, and supports accurate overcharge risk assessment and stable system black start. It also solves problems such as power outages and black start failures caused by hard-cutoff protection in existing technologies.

Description

Technical Field

[0001] This invention belongs to the field of lithium battery safety control technology, specifically relating to a protection and control method for overcharging of generator lithium batteries. Background Technology

[0002] With the upgrading of industrial automation and emergency power supply needs, generator sets integrating intelligent monitoring and hybrid power supply are becoming increasingly popular. Lithium batteries have transformed from auxiliary starting power supplies into core components that carry starting, control and transient buffering functions. The complexity of operating conditions has increased significantly, and they are susceptible to transient events such as generator sudden unloading and excitation hysteresis, which can lead to overcharging risks. Especially in unattended scenarios, this can easily lead to a chain of failures.

[0003] Currently, industry-standard protection measures for such scenarios still rely on independent module-based control. On the lithium battery side, the battery management system monitors parameters and triggers threshold protection. On the generator side, an automatic voltage regulator provides basic charging control, with some scenarios using simple additional devices to handle voltage spikes. Overall, existing technologies achieve some degree of post-overcharge risk prevention, but the system only activates localized protection for individual modules when battery parameters approach thresholds and a risk has already materialized. Essentially, it's a remedy for existing damage rather than proactive prevention.

[0004] Therefore, it can be seen that the protection of existing technical solutions is based on the premise that the battery state reaches a threshold, lacking a mechanism for predicting transient risks and regulating energy. The overcharge risk in the generator scenario is not caused by a single factor. The generator overcharge risk often stems from the superposition of three factors: the rise in bus voltage after sudden unloading, the residual excess energy due to the lag in excitation regulation, and the inertial continuation of the charging branch current. When the battery management system detects that the voltage is approaching the upper limit, transient overcharge energy has already formed and begun to be conducted to the battery. At this time, executing cut-off or current limiting is very likely to cause cell overvoltage damage due to the lag in action. At the same time, the existing technology has not established a cross-module collaborative control system, treating battery protection and generator control separately. When the battery management system hard disconnects the charging circuit or the contactor is forcibly disconnected, although the battery overcharge is prevented, the control power supply will be lost, which will lead to the failure of the next black start, and cannot meet the requirements for long-term reliable operation of unattended emergency generator sets. In addition, the protection actions of the existing technology are all single hard cut-off modes, lacking targeted hierarchical protection strategies. Executing overly aggressive protection actions in low-risk scenarios will affect the continuity of system operation and further reduce the reliability of the system. In view of this, the present invention proposes a protection and control method for overcharging of generator lithium batteries. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] This invention provides a protection and control method for overcharging a generator lithium battery. The purpose is to solve the problem that the control power fails when the charging circuit is hard disconnected or the contactor is forcibly disconnected in the existing generator lithium battery charging system, resulting in the failure of the next black start. Therefore, this invention provides a protection and control method for overcharging a generator lithium battery.

[0007] To at least solve the above problems, the present invention provides the following technical solution: A method for overcharge protection and control of a generator lithium battery includes: Real-time status data from the generator side and the lithium battery side are acquired to obtain a joint monitoring dataset; Event identification is performed on the joint monitoring dataset to obtain overcharge trigger candidate events; The battery approach state is determined based on the joint monitoring dataset, and the overcharge risk assessment result is obtained based on the overcharge trigger candidate events and the battery approach state. Based on the overcharge risk assessment results, the protection level is matched to obtain the protection level determination result; The protection level determination results are analyzed to obtain a set of protection parameters.

[0008] As a preferred implementation, event identification is performed on the joint monitoring dataset to obtain overcharge trigger candidate events, including: The joint monitoring dataset includes overcharge-related events. A preset feature recognition window is used to identify and extract the feature parameters of the overcharge-related events to obtain feature recognition results. The preset event judgment threshold is used to determine the deviation range of the feature recognition results, and events that meet the preset triggering conditions are judged as overcharge triggering candidate events.

[0009] As a preferred implementation, the feature parameters of the overcharge-related event are identified and extracted to obtain the feature identification result, including: The runtime data of the overcharge-related event is input into the feature recognition window for sampling and parsing. The parsed parameters are identified and extracted to obtain the event runtime features of the overcharge-related event. The event's operational characteristics are categorized and determined to obtain feature recognition results.

[0010] As a preferred implementation, the overcharge risk assessment result is obtained based on the overcharge trigger candidate events and the battery approach state, including: Based on preset quantization rules, feature extraction and quantization assignment are performed on the overcharge trigger candidate events to obtain overcharge trigger quantization values; The characteristic parameters of the battery approximation state are fused and processed. The numerical superposition and normalization operations of the characteristic parameters are performed according to the preset weights to obtain the risk warning value. The overcharge trigger quantization value and the risk warning value are fused, calculated, and weighted to obtain the fused risk value; The risk level of the fusion risk value is calibrated and its validity is verified to obtain the overcharge risk assessment result.

[0011] As a preferred implementation, the fusion risk value is calibrated for risk level and its validity is verified to obtain an overcharge risk assessment result, including: The fusion risk value is matched and compared based on a preset risk threshold range, and the risk level is calibrated according to the range in which the fusion risk value is located, to obtain a preliminary risk level result; Retrieve real-time operational data from the joint monitoring dataset, and compare the preliminary risk level results with the real-time operational data to obtain valid calibration results; The effective calibration results are then quantified and assigned values ​​to obtain the quantitative assessment results of overcharging risk.

[0012] As a preferred implementation, protection level matching is performed based on the overcharge risk assessment results to obtain a protection level determination result, including: Based on the preset risk protection level correspondence, the overcharge risk assessment results are matched and associated to obtain the corresponding risk level information; The risk level information is parsed to obtain level identification information; the protection execution strategy and benchmark control parameters are retrieved based on the level identification information, and the protection execution strategy and benchmark control parameters are associated and bound with the level identification information to generate a protection level determination result.

[0013] As a preferred implementation, the overcharge risk assessment results are matched and correlated to obtain corresponding risk level information, including: The overcharge risk quantification assessment result is compared with the preset risk level range to obtain the risk level labeling result; The risk level labeling results are adapted and matched based on preset mapping rules to obtain the adapted protection level. The adaptation protection level is associated with the overcharge risk quantification assessment result to obtain the corresponding risk level information.

[0014] As a preferred implementation, the risk level information is parsed to obtain level identification information, including: The risk level information is analyzed for numerical dimensions and level features to extract numerical dimensions and level features; The numerical dimension is compared with the preset level determination threshold to determine the current risk level identifier and obtain the level identifier information.

[0015] In a preferred embodiment, the numerical dimension is compared with a preset level determination threshold to determine the current risk level identifier, thereby obtaining level identifier information, including: The numerical dimensions are analyzed step by step against preset multi-level judgment thresholds to obtain the corresponding risk level range; The risk level range is validated, and the target risk level range is locked based on the numerical dimension after the validity validation, and a risk level identifier is generated. The risk level identifiers are organized and summarized to obtain the level identifier information.

[0016] As a preferred implementation, a closed-loop analysis is performed on the protection level determination result to obtain a protection parameter set, including: The protection execution strategy and baseline control parameters in the protection level determination result are retrieved to obtain the baseline control parameters to be executed; The reference control parameters to be executed are sent out for execution, and real-time feedback status data from the generator side and the lithium battery side are collected to obtain real-time operation feedback data. The real-time operation feedback data and preset safety constraints are dynamically corrected to obtain preliminary corrected control parameters; The preliminary control parameters are adjusted in steps with limiting until the real-time operating status of the generator side and the lithium battery side meets the preset safety constraints, thus forming the final protection parameter set.

[0017] Compared with the prior art, the present invention has the following advantages: 1. By synchronously collecting real-time status data from the generator and lithium battery sides to construct a joint monitoring dataset, and combining overcharge trigger candidate events with battery approach states for fusion calculation, transient risks can be identified and overcharge risk assessment results can be generated before overcharge actually occurs. This provides a proactive prediction function for transient risks, realizing the transformation from remediation after threshold triggering to pre-control of risk precursors. It effectively avoids the problem of energy conduction lag caused by sudden generator unloading, excitation lag, and charging inertia superposition, significantly reducing the probability of cell overvoltage damage and improving the operational safety of lithium batteries under complex transient conditions. Furthermore, a weighted fusion method of overcharge trigger quantification value and risk precursor value is adopted to obtain a fused risk value. Through multi-level threshold interval calibration, real-time operation data verification, and validity validation, a quantitative and reliable overcharge risk assessment result is formed. Compared with the single threshold judgment method, it greatly improves the risk identification accuracy and anti-interference capability, and can effectively filter out misjudgments caused by abnormal disturbances and instantaneous fluctuations. It is especially suitable for unattended emergency generator set scenarios with variable operating conditions, ensuring the reliability and consistency of protection actions.

[0018] 2. This invention, based on a risk level classification and matching mechanism, correlates the quantitative assessment results of overcharge risk with protection level, protection execution strategy, and control parameters. This achieves refined and graded linkage between risk level and protection action, avoiding one-size-fits-all protection control. It enables flexible adjustment in low-risk areas and rapid response in high-risk areas, suppressing overcharge while maximizing normal system operation, achieving energy regulation, and improving the rationality and adaptability of overall control. By performing closed-loop analysis on the protection level determination results, and using real-time operational feedback data as a basis, the baseline control parameters are dynamically corrected and adjusted step by step to form the final protection parameter set. This achieves coordinated control and closed-loop optimization between the generator side and the battery side, breaking the traditional problem of disconnection between battery protection and generator control. It avoids power outages caused by hard disconnection of the charging circuit or forced disconnection of the contactor, fundamentally solving cascading faults such as black start failure and system power failure shutdown, and improving the continuous operation capability of the emergency power supply system.

[0019] 3. This invention introduces quantitative processing methods such as numerical normalization, weight calculation, multi-level threshold determination, and validity verification throughout the entire protection and control process. This makes risk assessment, level determination, and parameter adjustment reproducible, calibrable, and iterative. It can adapt to generators of different capacities, lithium battery packs of different types, and different operating conditions. It can complete adaptive adjustment without relying on manual intervention, reduce on-site operation and maintenance pressure, and improve the intelligence level and long-term stability of unattended emergency power supply systems. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the process of the present invention.

[0021] Figure 2 This is a comparison curve of generator bus voltage and lithium battery cell voltage under different protection schemes. Detailed Implementation

[0022] To make the technical means, creative features, and achieved objectives and effects of this invention easier to understand, the invention is further described below with reference to specific embodiments. However, the following embodiments are merely preferred embodiments of this invention and not all of them. Other embodiments obtained by those skilled in the art based on the embodiments described herein without creative effort are all within the protection scope of this invention. Unless otherwise specified, the experimental methods in the following embodiments are conventional methods, and the materials and reagents used in the following embodiments are commercially available unless otherwise specified.

[0023] Combination Figure 1 The illustrated exemplary flowchart of a generator lithium battery overcharge protection control method includes the following implementation steps: Real-time status data from the generator side and the lithium battery side are acquired to obtain a joint monitoring dataset; This step is used to collect real-time operational data from the generator side, lithium battery side, and charging branch, and to build a unified joint monitoring dataset. This provides comprehensive and accurate data input for subsequent overcharge trigger event identification, risk assessment, and protection control, breaking the problem of data separation between generator control and battery protection in existing technologies. The aforementioned joint monitoring data is collected synchronously from the generator side, lithium battery side, and charging branch through a preset sampling module, sensor, and controller. After data preprocessing, a structured dataset is formed, which provides unified data support for the identification of overcharge trigger events, the determination of battery near-state conditions, and risk assessment.

[0024] Specifically, real-time operating parameters of the generator side, real-time status parameters of the lithium battery side, and real-time working parameters of the charging branch are collected to complete the full-dimensional data collection. After noise reduction and normalization preprocessing, a joint monitoring dataset is obtained. Among them, the real-time operating parameters on the generator side are obtained from the generator controller, automatic voltage regulator and speed sensor, including generator output voltage, generator speed, excitation current, excitation duty cycle, generator load current and bus voltage. The generator output voltage is used to reflect the stability of generator power supply, the generator speed is used to monitor the speed recovery after sudden unloading, the excitation current and excitation duty cycle are used to characterize the magnitude of generator excitation energy, the generator load current is used to identify load drop events, and the bus voltage is used to monitor the voltage rise after sudden unloading. The generator controller is a core control module used to monitor and control the generator's operating status in real time. It is pre-integrated into the emergency generator set and is used to output operating parameters such as generator output voltage, speed, and excitation current. The automatic voltage regulator is used to adjust the generator excitation current to maintain a stable generator output voltage and synchronously output parameters such as excitation current and excitation duty cycle. The speed sensor is used to collect the generator rotor speed in real time, output the speed signal, and convert it into an acquireable electrical signal. The real-time status parameters of the lithium battery side come from the battery management system and temperature sensor, including the total voltage of the lithium battery pack, the highest single-cell voltage, the lowest single-cell voltage, the battery temperature, the charging current, and the status word of the battery management system. The total voltage of the lithium battery pack and the highest single-cell voltage are used to determine the degree to which the battery is approaching overcharge, the lowest single-cell voltage is used to help determine the battery consistency, the battery temperature is used to monitor temperature anomalies during overcharge, the charging current is used to reflect the charging intensity, and the status word of the battery management system is used to identify the battery management system's warning, hard protection, and other states. The battery management system is a dedicated control module for monitoring the operating status of lithium batteries and protecting their safety. It is pre-integrated into the lithium battery pack and outputs the total battery voltage, individual cell voltage, temperature, charging current, and its own operating status. The temperature sensors are installed on the surface of the lithium battery cells and inside the battery box to collect battery temperature in real time and prevent overcharging hazards caused by excessive temperature. The real-time operating parameters of the charging branch are obtained from the Hall current sensor and voltage sampling module, including the charging branch current, charging branch voltage and rectifier firing angle, which are used to reflect the operating status of the charging branch and identify abnormal situations such as continuous conduction of the charging branch and current inertia continuation. The aforementioned joint monitoring dataset refers to a structured dataset that, after collection and preprocessing, includes all real-time operating parameters from the generator side, lithium battery side, and charging branch. Furthermore, the collected raw data is preprocessed, specifically including: using the moving average filtering method to reduce noise in parameters with large fluctuations such as generator speed, charging current, and battery voltage, and eliminating instantaneous interference data; normalizing parameters of different dimensions, mapping the values ​​of each parameter to the [0,1] interval, and eliminating the impact of dimensional differences on subsequent fusion calculations; supplementing missing data using linear interpolation; and marking and temporarily storing abnormal data that exceeds the preset reasonable range, which will not be included in subsequent calculations and will be re-verified after the data is supplemented. Furthermore, the data acquisition cycle is configured, with different acquisition cycles set according to the complexity of the operating conditions: for unattended operation and high-risk sudden unloading scenarios, a short acquisition cycle is set, with a cycle range of 5ms to 10ms, to quickly capture transient events such as sudden unloading and speed recovery; for normal stable operation scenarios, a medium acquisition cycle is set, with a cycle range of 20ms to 50ms, to balance the real-time performance of data acquisition with system resource consumption; for low-load and low-risk scenarios, a long acquisition cycle is set, with a cycle range of 100ms to 200ms, to reduce system energy consumption. The scenarios described above, where there is a high risk of unattended operation and sudden unloading, refer to situations where the emergency generator set is operating at full load, frequently switching loads, or where the surrounding load fluctuates significantly. The system automatically determines the load current fluctuation based on the magnitude of the fluctuation. When the load current fluctuation exceeds a preset fluctuation threshold of ±20%, it automatically switches to short-cycle data acquisition. The preset fluctuation threshold is pre-calibrated based on the generator set's rated power and load type to adapt to the operating characteristics of different models.

[0025] Event identification is performed on the joint monitoring dataset to obtain overcharge trigger candidate events; The event identification mentioned above refers to the process of extracting features and comparing thresholds of overcharge-related events in the joint monitoring dataset based on a preset feature identification window to determine whether they are high-risk transient events. This step is performed to screen out candidate events that may cause overcharging and provide a triggering basis for subsequent risk assessment.

[0026] Specifically, the joint monitoring dataset includes overcharge-related events and various operational events related to overcharge risk. A preset feature recognition window is used to identify and extract feature parameters of overcharge-related events to obtain feature recognition results. A preset event judgment threshold is used to determine the deviation range of the feature recognition results, and events that meet the preset triggering conditions are judged as overcharge triggering candidate events. The overcharge-related events include load descent events, speed recovery events, bus rise events, battery management system warning events, and charging branch continuous conduction events. Specifically, the characteristic parameter of the load descent event is the rate of change of the generator load current. The preset feature recognition window is 100ms, that is, 10 short-cycle data are continuously collected, and the rate of change of the load current within 100ms is extracted to obtain the feature recognition result. The preset event judgment threshold is a load current drop of ≥30%. Relative to the current steady-state load current, when the load current drop in the feature recognition result meets this threshold and the duration is ≥50ms, it is judged as a load descent event and included in the overcharge trigger candidate event. The characteristic parameter of the speed recovery event is the generator speed deviation. The preset feature recognition window is 200ms. The speed deviation within 200ms is continuously collected and extracted to obtain the feature recognition result. The preset event judgment threshold is a speed deviation ≥5%. When the speed deviation in the feature recognition result meets this threshold and the duration is ≥100ms, it is judged as a speed recovery event and included in the overcharge trigger candidate event. The steady-state reference speed is the speed value within ±1% of the rated speed of the generator set, which is calibrated by the generator set factory parameters. The characteristic parameter of the bus rise event is the DC bus voltage rise slope. The preset feature recognition window is 50ms. The change in bus voltage within 50ms is calculated to obtain the voltage rise slope. The preset event judgment threshold is a voltage rise slope ≥ 0.5V / ms. When the feature recognition result meets this threshold, it is judged as a bus rise event and included in the overcharge trigger candidate event. The characteristic parameter of the battery management system pre-alarm event is the battery management system status word. The preset feature recognition window is one collection cycle, and the battery management system status word is extracted. When the battery management system status word contains any of the identifiers "cell voltage approach warning", "temperature warning" or "charging current over-limit warning", it is determined to be a battery management system pre-alarm event and included in the overcharge trigger candidate event. The battery management system status word is generated by the battery management system according to the preset coding rules. Different warning types correspond to different status codes and can be directly extracted from the joint monitoring dataset. The characteristic parameter of the continuous conduction event of the charging branch is the charging branch current. The preset feature recognition window is 500ms. The charging branch current within 500ms is continuously collected and extracted. The preset event judgment threshold is that the charging branch current is ≥0.5A. When the feature recognition result meets this threshold and the duration is ≥300ms, it is judged as a continuous conduction event of the charging branch and included in the overcharge trigger candidate event. The feature recognition window refers to the time window used for continuously collecting and parsing the feature parameters of overcharge-related events. The feature recognition window for different overcharge-related events is set according to the event response speed and duration to ensure the completeness and accuracy of feature parameter extraction. Among them, the feature recognition result refers to the set of feature parameters extracted from the overcharge-related event that can characterize the severity of the event. It is generated by the feature extraction step and is used to compare with the preset event judgment threshold to determine whether it is an overcharge trigger candidate event. Furthermore, the feature parameters of the overcharge-related event are identified and extracted to obtain the feature identification result. Specifically, this includes: inputting the operation data of the overcharge-related event into the feature identification window for sampling and parsing; filtering, identifying and extracting the parsed parameters to obtain the event operation features of the overcharge-related event; classifying and judging the event operation features according to the event type, eliminating invalid features, and retaining the valid features that can characterize the event risk to obtain the feature identification result. Furthermore, when multiple overcharge trigger candidate events are identified simultaneously, the events are prioritized and assigned different priorities according to their impact on overcharge risk: battery management system warning events have the highest priority, as they are directly related to the battery's approach state; followed by load drop events and speed recovery events; and finally, bus rise events and charging branch continuous conduction events. In subsequent risk assessments, different quantitative weights will be assigned based on the event priorities to improve the accuracy of risk assessments.

[0027] The battery approach state is determined based on the joint monitoring dataset, and the overcharge risk assessment result is obtained based on the overcharge trigger candidate events and the battery approach state. This step is used to combine the joint monitoring dataset to quantitatively determine the degree to which the battery is approaching overcharge, and at the same time integrate the risk levels of candidate overcharge triggering events. The overcharge risk assessment result is obtained through weighted fusion calculation, realizing the quantitative and accurate assessment of overcharge risk. This is different from the qualitative judgment method of a single threshold in the existing technology, and provides a quantitative basis for subsequent graded protection.

[0028] The overcharge risk assessment refers to the process of obtaining a quantitative and reliable overcharge risk assessment result by combining the quantitative value of the overcharge trigger candidate event and the risk precursor value of the battery approaching the state through fusion calculation, level calibration and validity verification. This is used to transform qualitative event identification into quantitative risk assessment to support the accurate matching of subsequent protection levels.

[0029] Specifically, the battery approach state is first determined based on the joint monitoring dataset. The battery approach state refers to the degree to which the current operating state of the lithium battery is close to the overcharge threshold. It is determined by the lithium battery side parameters in the joint monitoring dataset: total voltage of the lithium battery pack, highest single cell voltage, battery temperature, and charging current. The rules for determining the battery's near-overcharge state are as follows: Read the highest single-cell voltage Vmax, single-cell warning voltage Vpre, and single-cell protection upper limit voltage Vlim from the joint monitoring dataset. Calculate the margin Vlim-Vmax between the highest single-cell voltage and the warning threshold. The smaller the margin, the higher the degree of battery nearing overcharge. The single-cell warning voltage Vpre and single-cell protection upper limit voltage Vlim are provided by the cell manufacturer. Read the battery temperature T. When T ≥ 55℃, it is determined to be an abnormal temperature approach, exacerbating the risk of overcharge. Read the charging current Ichg. When Ichg ≥ Iallow (allowable charging current), it is determined to be an abnormal charging current approach, accelerating the overcharge process. Based on the above determination results, the characteristic parameters of the battery approach state are determined, including: the maximum single-cell voltage margin, the maximum single-cell voltage rise rate dVmax / dt, the normalized value of charging current Ichg / Iallow, and the normalized value of battery temperature T / Tlim, where Tlim is the maximum allowable temperature of the battery, preferably 60℃. Furthermore, this embodiment proposes to obtain an overcharge risk assessment result based on the overcharge trigger candidate event and the battery approach state, specifically including the following steps: Based on preset quantization rules, feature extraction and quantization assignment are performed on the overcharge trigger candidate events to obtain overcharge trigger quantization values; The aforementioned preset quantification rules refer to the rules for assigning different quantification scores based on the type and severity of overcharge trigger candidate events, combined with event priority and impact. Specific quantification scoring standards are as follows: Battery management system warning events are assigned 8-10 points, with different scores for different warning types: single cell voltage approach warning 10 points, temperature warning 9 points, charging current over-limit warning 8 points; load drop events are assigned 6-8 points, with higher scores for larger drops: ≥50% drop is 8 points. 30%-50% is worth 6 points; speed recovery event is worth 5-7 points, with higher scores for larger speed deviations: deviation ≥10% is worth 7 points, 5%-10% is worth 5 points; bus rise event is worth 4-6 points, with higher scores for larger voltage rise slopes: ≥1.0V / ms is worth 6 points, 0.5-1.0V / ms is worth 4 points; charging branch continuous conduction event is worth 3-5 points, with higher scores for longer conduction time: ≥1000ms is worth 5 points, 300-1000ms is worth 3 points; For each identified overcharge trigger candidate event, a quantification value is assigned according to the above criteria. Then, the quantification scores of all events are summed to obtain the overcharge trigger quantification value Trig. The Trig value ranges from 0 to 40 points. The higher the score, the higher the risk of overcharge triggering. If no overcharge trigger candidate event is identified, Trig = 0 points, and it is determined that there is no risk of overcharge triggering. The feature parameters of the battery approximation state are fused and processed. The feature parameters are then subjected to numerical superposition and normalization operations according to preset weights to obtain the risk precursor value Roc. The aforementioned risk precursor value (Roc) is a core quantitative indicator used to characterize the degree to which a battery approaches overcharge. It can reflect the development trend of overcharge risk before the battery reaches the overcharge threshold. Its calculation formula is as follows: ; in, The highest voltage currently available for a single cell; Individual warning voltage; Individual unit protection upper limit voltage; : Maximum single-unit voltage rise rate; : Normalized reference value for voltage rise rate; Current charging current; Allowable charging current; Current battery temperature; Maximum permissible temperature of the battery; .

Claims

1. A method for overcharging protection and control of a generator lithium battery, characterized in that, include: Real-time status data from the generator side and the lithium battery side are acquired to obtain a joint monitoring dataset; Event identification is performed on the joint monitoring dataset to obtain overcharge trigger candidate events; The battery approach state is determined based on the joint monitoring dataset, and the overcharge risk assessment result is obtained based on the overcharge trigger candidate events and the battery approach state. Based on the overcharge risk assessment results, the protection level is matched to obtain the protection level determination result; The protection level determination results are analyzed to obtain a set of protection parameters.

2. The protection and control method according to claim 1, characterized in that, Event identification is performed on the joint monitoring dataset to obtain overcharge trigger candidate events, including: The joint monitoring dataset includes overcharge-related events. A preset feature recognition window is used to identify and extract the feature parameters of the overcharge-related events to obtain feature recognition results. The preset event judgment threshold is used to determine the deviation range of the feature recognition results, and events that meet the preset triggering conditions are judged as overcharge triggering candidate events.

3. The protection and control method according to claim 2, characterized in that, The feature parameters of overcharge-related events are identified and extracted to obtain feature identification results, including: The runtime data of the overcharge-related event is input into the feature recognition window for sampling and parsing. The parsed parameters are identified and extracted to obtain the event runtime features of the overcharge-related event. The event's operational characteristics are categorized and determined to obtain feature recognition results.

4. The protection and control method according to claim 1, characterized in that, The overcharge risk assessment results are obtained based on the overcharge trigger candidate events and the battery approach state, including: Based on preset quantization rules, feature extraction and quantization assignment are performed on the overcharge trigger candidate events to obtain overcharge trigger quantization values; The characteristic parameters of the battery approximation state are fused and processed. The numerical superposition and normalization operations of the characteristic parameters are performed according to the preset weights to obtain the risk warning value. The overcharge trigger quantization value and the risk warning value are fused, calculated, and weighted to obtain the fused risk value; The risk level of the fusion risk value is calibrated and its validity is verified to obtain the overcharge risk assessment result.

5. The protection and control method according to claim 4, characterized in that, The fusion risk value is calibrated for risk level and its validity is verified to obtain the overfill risk assessment result, including: The fusion risk value is matched and compared based on a preset risk threshold range, and the risk level is calibrated according to the range in which the fusion risk value is located, to obtain a preliminary risk level result; Retrieve real-time operational data from the joint monitoring dataset, and compare the preliminary risk level results with the real-time operational data to obtain valid calibration results; The effective calibration results are then quantified and assigned values ​​to obtain the quantitative assessment results of overcharging risk.

6. The protection and control method according to claim 1, characterized in that, Based on the overcharge risk assessment results, protection level matching is performed to obtain the protection level determination result, including: Based on the preset risk protection level correspondence, the overcharge risk assessment results are matched and associated to obtain the corresponding risk level information; The risk level information is parsed to obtain level identification information; the protection execution strategy and benchmark control parameters are retrieved based on the level identification information, and the protection execution strategy and benchmark control parameters are associated and bound with the level identification information to generate a protection level determination result.

7. The protection and control method according to claim 6, characterized in that, The overcharge risk assessment results are matched and correlated to obtain the corresponding risk level information, including: The overcharge risk quantification assessment result is compared with the preset risk level range to obtain the risk level labeling result; The risk level labeling results are adapted and matched based on preset mapping rules to obtain the adapted protection level. The adaptation protection level is associated with the overcharge risk quantification assessment result to obtain the corresponding risk level information.

8. The protection and control method according to claim 6, characterized in that, The risk level information is parsed to obtain level identification information, including: The risk level information is analyzed for numerical dimensions and level features to extract numerical dimensions and level features; The numerical dimension is compared with the preset level determination threshold to determine the current risk level identifier and obtain the level identifier information.

9. The protection and control method according to claim 8, characterized in that, The numerical dimension is compared with a preset level determination threshold to determine the current risk level identifier, thus obtaining level identifier information, including: The numerical dimensions are analyzed step by step against preset multi-level judgment thresholds to obtain the corresponding risk level range; The risk level range is validated, and the target risk level range is locked based on the numerical dimension after the validity validation, and a risk level identifier is generated. The risk level identifiers are organized and summarized to obtain the level identifier information.

10. The protection and control method according to claim 1, characterized in that, A closed-loop analysis is performed on the protection level determination results to obtain a set of protection parameters, including: The protection execution strategy and baseline control parameters in the protection level determination result are retrieved to obtain the baseline control parameters to be executed; The reference control parameters to be executed are sent out for execution, and real-time feedback status data from the generator side and the lithium battery side are collected to obtain real-time operation feedback data. The real-time operation feedback data and preset safety constraints are dynamically corrected to obtain preliminary corrected control parameters; The preliminary control parameters are adjusted in steps with limiting until the real-time operating status of the generator side and the lithium battery side meets the preset safety constraints, thus forming the final protection parameter set.

Images