Battery safety protection system based on multi-scale flame-retardant network

Abstract

The invention discloses a battery safety protection system based on a multi-scale flame-retardant network. The system comprises a multi-scale flame-retardant network construction module, a battery operation parameter sensing fusion module, a battery safety state evaluation module, a risk early warning decision module, a flame-retardant network dynamic regulation and control module and a cooperative linkage external control module. According to the system, a three-dimensional flame-retardant network with a fractal structure is constructed, multi-dimensional operation parameters are collected and fused in real time, the safety state of a battery is evaluated based on a dynamic threshold value, a response instruction is generated according to a risk level, and the flame-retardant network is dynamically adjusted and linked with an external system. All the modules operate cooperatively, comprehensive monitoring and precise protection of the safety state of the battery are achieved, the occurrence probability of battery safety accidents is effectively reduced, and the safety and reliability of battery operation are improved.

Description

Technical Field

[0001] This invention relates to the field of battery safety protection, and more particularly to a battery safety protection system based on a multi-scale flame-retardant network. Background Technology

[0002] With the rapid development of new energy technologies, the safety of batteries, as core energy storage components, has become an increasingly important concern. In fields such as electric vehicles and energy storage power stations, fires and explosions caused by overcharging, overheating, and short circuits are frequent, resulting not only in significant property damage but also threatening human lives. Traditional battery safety protection methods, such as simple fire-resistant materials and basic temperature monitoring devices, are no longer sufficient to meet the demands of increasingly complex application scenarios.

[0003] Existing battery safety protection technologies have several shortcomings. First, the protection system lacks systematicity and multi-scale coordination. Most existing solutions only protect a single aspect of battery safety, such as simply relying on flame-retardant materials to wrap the battery or using only temperature sensors for risk warnings. There is a lack of effective data interaction and linkage mechanisms between the various protection units, failing to form a comprehensive and efficient protection network. When faced with complex and ever-changing battery safety risks, the protection effect is significantly reduced. Second, the assessment and response to battery safety status lacks dynamism and precision. Traditional assessment methods often rely on fixed thresholds to determine battery safety status, failing to adapt to changes in battery operating conditions and environmental factors, easily leading to misjudgments or omissions. In terms of risk response, it is also difficult to implement graded and precise control strategies, resulting in the inability to take timely and effective measures to contain and handle battery safety incidents. Therefore, an innovative battery safety protection system is urgently needed to improve battery safety performance. Summary of the Invention

[0004] To overcome the shortcomings and deficiencies of existing technologies, this invention provides a battery safety protection system based on a multi-scale flame-retardant network.

[0005] The technical solution adopted in this invention is a battery safety protection system based on a multi-scale flame-retardant network, comprising:

[0006] The multi-scale flame retardant network construction module is used to construct a multi-scale flame retardant network with a fractal structure based on the spatial distribution characteristics of the battery pack. Through the nested arrangement of flame retardant units at different levels in three-dimensional space, a three-dimensional flame retardant protection system is formed, and bidirectional data interaction is performed with the battery safety protection analysis module.

[0007] The battery operating parameter sensing and fusion module collects multi-dimensional operating parameters of battery voltage, current, temperature and pressure in real time through an integrated sensor array, performs spatiotemporal alignment and feature fusion processing on the collected data, and transmits the processed data to the battery safety status assessment module.

[0008] The battery safety status assessment module receives data from the battery operation parameter sensing and fusion module, maps real-time parameters to corresponding safety status intervals through a preset multi-scale safety threshold matrix, generates battery safety status assessment results, and sends them to the risk warning decision module.

[0009] The risk warning and decision-making module generates corresponding risk warnings and control commands based on the battery safety status assessment results and a pre-built risk level-response strategy mapping table. These commands are then transmitted to the flame retardant network control module and the external linkage control module, respectively.

[0010] The flame retardant network dynamic control module receives control commands from the risk warning decision module and, in conjunction with the optimization parameters output by the battery safety protection analysis module, dynamically adjusts the flame retardant release amount and flame retardant structural morphology of the multi-scale flame retardant network.

[0011] The collaborative external control module receives control commands from the risk warning and decision-making module and communicates with the fire protection system, ventilation system, and power management system to perform multi-system collaborative control.

[0012] The multi-scale flame retardant network construction module, battery operating parameter sensing and fusion module, battery safety status assessment module, risk early warning decision module, flame retardant network dynamic control module, and collaborative external control module are connected in sequence to form a complete battery safety protection system.

[0013] Furthermore, the multi-scale flame-retardant network construction module adopts the following optimization model formula when constructing the multi-scale flame-retardant network:

[0014]

[0015] in, This indicates the comprehensive protective effectiveness of a multi-scale flame-retardant network; This is a global adjustment coefficient, ranging from 0.8 to 1.2, used to balance the influence weight of each level of flame-retardant unit as a whole; The number of layers in the flame-retardant network; For the first The weighting coefficient of each level is determined by the complexity of the battery pack's spatial layout; For the first The topological parameters of the tiered flame-retardant unit describe its spatial arrangement; For the first The effective working volume of the tiered flame-retardant unit; For the first The flame propagation suppression parameters of the tiered flame retardant unit.

[0016] Furthermore, the battery operating parameter sensing and fusion module employs the following optimized model formula when performing feature fusion processing on the collected data:

[0017]

[0018] in, This represents the feature vector of the fused data; The fusion weighting coefficient has a value range between 0.7 and 1.3. Number of sensor types; For the first The weight of sensor data is determined by the importance of the sensor in battery safety monitoring; For the first A vector obtained after normalization of sensor data; This is the trend weighting coefficient; This is a time series trend feature vector for multi-dimensional data.

[0019] Furthermore, the battery safety status assessment module uses the following optimized model formula when performing safety status assessment:

[0020]

[0021] in, This indicates the battery safety status assessment value; To evaluate the number of different types of parameters; For the first The weight of each evaluation parameter is determined by its degree of impact on battery safety; Indicates the first Real-time parameters Mapped to the corresponding security threshold range The mapping function outputs the corresponding security status score.

[0022] Furthermore, the risk warning decision module uses the following optimized model formula when generating risk warning and control instructions:

[0023]

[0024] in, This represents the generated risk response instruction vector; Number of risk levels; For the first Weighting coefficients for each risk level; This indicates the battery safety status assessment value. With the Risk Level The matching function outputs the response strategy parameters corresponding to the risk level.

[0025] Furthermore, the flame-retardant network dynamic control module uses the following optimization model formula when dynamically adjusting the flame-retardant network:

[0026]

[0027] in, This represents the adjustment parameter vector of the flame-retardant network; The number of types of adjustable parameters; For the first The weighting coefficients of the adjustment parameters; Adjust Indicates according to control commands Optimization parameters output by the battery safety protection analysis module For the A function that adjusts various parameters.

[0028] Furthermore, the battery safety status assessment module includes: a voltage parameter mapping assessment unit, which receives battery voltage data transmitted by the battery operating parameter sensing and fusion module, maps real-time voltage parameters to corresponding safety status intervals according to a preset voltage safety threshold matrix, and generates a voltage safety status assessment result; a current parameter mapping assessment unit, which receives battery current data, maps real-time current parameters to corresponding safety status intervals according to a preset current safety threshold matrix, and outputs a current safety status assessment result; a temperature parameter mapping assessment unit, which receives battery temperature data, maps real-time temperature parameters to corresponding safety status intervals according to a preset temperature safety threshold matrix, and obtains a temperature safety status assessment result; and a pressure parameter mapping assessment unit, which receives battery pressure data, maps real-time pressure parameters to corresponding safety status intervals according to a preset pressure safety threshold matrix, and obtains a pressure safety status assessment result.

[0029] Furthermore, the risk warning decision module includes: a low-risk warning instruction generation unit, which generates corresponding low-risk warning information and preliminary control instructions based on a pre-built risk level-response strategy mapping table when the battery safety status assessment result is at a low-risk level; a medium-risk warning instruction generation unit, which generates medium-level risk warning information and enhanced control instructions based on a pre-built risk level-response strategy mapping table when the battery safety status assessment result is at a medium-risk level; a high-risk warning instruction generation unit, which generates high-risk warning information and emergency control instructions based on a pre-built risk level-response strategy mapping table when the battery safety status assessment result is at a high-risk level; and a risk instruction priority determination unit, which prioritizes the generated warning and control instructions of different risk levels to ensure that high-priority instructions are executed first when multiple risks coexist.

[0030] Furthermore, the flame retardant network dynamic control module includes: a flame retardant release amount precision adjustment unit, which receives control commands from the risk warning decision module and, in conjunction with optimization parameters output by the battery safety protection analysis module, precisely adjusts the release rate and amount of flame retardant in the multi-scale flame retardant network; a flame retardant structure morphology dynamic adjustment unit, which dynamically changes the structure morphology of the flame retardant network according to control commands and optimization parameters, including adjusting the thickness and density parameters of the flame retardant layer; a flame retardant network local reinforcement unit, which, when a local safety hazard is detected in the battery, strengthens the flame retardant network in the corresponding area according to control commands to increase flame retardant protection capability; and a flame retardant network collaborative optimization unit, which coordinates parameter adjustments between various parts of the flame retardant network to ensure that the overall protective performance of the flame retardant network reaches its optimal level.

[0031] This application has the following beneficial effects:

[0032] This invention proposes a battery safety protection system based on a multi-scale flame-retardant network. This system constructs a multi-scale flame-retardant network based on topology adaptive adjustment, with flame-retardant units at each level arranged in a three-dimensional nested configuration to form a three-dimensional protection system. Through collaborative linkage with an external control module, it achieves deep communication with fire protection, ventilation, and other systems, enabling multi-dimensional and cross-system collaborative protection. Simultaneously, a battery operating parameter sensing and fusion module collects and processes data such as voltage, current, temperature, and pressure, with real-time data interaction between modules, constructing a complete and efficient protection closed loop. Addressing the shortcomings of traditional assessments in terms of inaccurate response and lack of dynamism, the battery safety status assessment module dynamically assesses the battery safety status based on a preset multi-scale safety threshold matrix and the fused multi-dimensional parameters, avoiding misjudgments and omissions caused by fixed thresholds. The risk warning and decision-making module accurately generates warnings and control commands corresponding to the risk level through a pre-constructed risk level-response strategy mapping table, achieving graded response. Furthermore, the flame-retardant network dynamic control module adjusts the release amount and structural morphology of the flame retardant in real time according to control commands and optimization parameters, further enhancing the real-time performance and effectiveness of protection, comprehensively improving the battery's safety performance under complex operating conditions, and reducing the risk of accidents. Attached Figure Description

[0033] Figure 1 This is a diagram showing the system module composition of the present invention;

[0034] Figure 2 This is a flowchart of the system operation of the present invention. Detailed Implementation

[0035] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0036] like Figure 1 As shown, a battery safety protection system based on a multi-scale flame-retardant network includes:

[0037] The multi-scale flame retardant network construction module is used to construct a multi-scale flame retardant network with a fractal structure based on the spatial distribution characteristics of the battery pack. Through the nested arrangement of flame retardant units at different levels in three-dimensional space, a three-dimensional flame retardant protection system is formed, and bidirectional data interaction is performed with the battery safety protection analysis module.

[0038] Specifically, the multi-scale flame-retardant network construction module aims to create a three-dimensional flame-retardant protection system with a fractal structure based on the spatial distribution characteristics of the battery pack. Through the nested arrangement of different levels of flame-retardant units, it achieves comprehensive flame-retardant protection for the battery pack. Simultaneously, it maintains bidirectional data interaction with the battery safety protection analysis module, ensuring that the flame-retardant network continuously optimizes based on battery safety analysis results. The multi-scale flame-retardant network constructed by this module can effectively suppress the spread of flames within the battery pack through the synergistic effect of different levels of flame-retardant units, enhancing the safety protection capability of the battery system. Its comprehensive protection effectiveness comprehensively considers the topology, effective volume, and flame-suppressing effect of each level of flame-retardant unit, providing fundamental protection for battery safety.

[0039] In its implementation, this module first conducts a detailed analysis of the battery pack's spatial layout to determine the number of layers in the flame-retardant network. Generally, the number of layers can be set to 3-5 layers, depending on the battery pack size and spatial complexity. The global adjustment coefficient is set between 0.8 and 1.2. When the battery pack's spatial layout is complex, a value biased towards 1.2 is used to balance the influence weight of each layer of flame-retardant units. The layer weight coefficient is determined by the complexity of the battery pack's spatial layout; the weight coefficient for the core area layer can be set to 0.3-0.5, while the weight coefficient for non-core area layers is set to 0.1-0.2. Simultaneously, the module precisely calculates the topological parameters of each layer of flame-retardant units to ensure full spatial coverage. The effective volume must be 1.2-1.5 times the size of the corresponding battery area. Flame suppression parameters are set based on the performance of the selected flame-retardant materials to ensure that each layer of flame-retardant units works synergistically to form an efficient three-dimensional flame-retardant protection system. The module also interacts with the battery safety protection analysis module in real time, adjusting network parameters based on the analysis results.

[0040] The battery operating parameter sensing and fusion module collects multi-dimensional operating parameters of battery voltage, current, temperature and pressure in real time through an integrated sensor array, performs spatiotemporal alignment and feature fusion processing on the collected data, and transmits the processed data to the battery safety status assessment module.

[0041] Specifically, the battery operating parameter sensing and fusion module primarily uses an integrated sensor array to collect multi-dimensional operating parameters such as battery voltage, current, temperature, and pressure in real time. The acquisition frequency is set to 10-20Hz to ensure timely capture of changes in the battery's operating status. This module also performs spatiotemporal alignment processing on the collected multi-dimensional data to eliminate data deviations caused by differences in acquisition time and location between different sensors. Then, feature fusion processing is performed to integrate the scattered single-parameter data into a data feature vector that comprehensively reflects the battery's operating status. This provides accurate and comprehensive data support for subsequent battery safety status assessments, avoiding inaccurate safety assessment results due to single or biased data.

[0042] In practical implementation, the integrated sensor array is arranged with a density of one voltage sensor for every 2-3 battery cells, one current sensor for every 5-8 battery cells, and one temperature and pressure sensor every 10-15 cm in key areas of the battery pack. After data acquisition, spatiotemporal alignment is performed first. Using the acquisition time of one core sensor as a benchmark, the data from other sensors are time-calibrated, with the deviation controlled within ±0.1s. Data matching is performed based on the sensor installation coordinates. During feature fusion processing, the fusion weight coefficient is set between 0.7 and 1.3, with a bias towards 1.3 when the battery is in a critical charging / discharging stage. Sensor data weights are set according to their importance: voltage and temperature sensor data weights are set to 0.3-0.4, current sensor data weights are set to 0.2-0.3, and pressure sensor data weights are set to 0.1-0.2. The data from various sensors are normalized to a range of 0-1. Then, the data is combined with a time series trend feature vector with a trend weight coefficient set to 0.2-0.3. By calculating the change trend of the data over the past 10-20 collection cycles, the fused data feature vector is finally generated and transmitted to the battery safety status assessment module.

[0043] The process of generating the fused data feature vector is based on multi-dimensional sensor data and is achieved through three steps: data preprocessing, weight allocation, and feature integration, as detailed below:

[0044] First, the raw data collected by the integrated sensor array, including voltage, current, temperature, and pressure, are preprocessed. First, spatiotemporal alignment is performed based on the acquisition time of the core sensor to eliminate data deviations caused by differences in acquisition time and installation location between different sensors (time calibration deviation is controlled within ±0.1s). Then, a normalization algorithm is used to map the data from various sensors to a unified numerical range of 0-1, eliminating dimensional differences in different parameter dimensions (such as voltage in V and temperature in °C), ensuring data comparability, and obtaining a standardized single-dimensional data sequence.

[0045] Secondly, weight coefficients are assigned to each sensor according to its importance in battery safety monitoring (e.g., voltage and temperature sensors are weighted at 0.3-0.4, current sensors at 0.2-0.3, and pressure sensors at 0.1-0.2). The standardized single-dimensional data are multiplied by their corresponding weights and summed to obtain the multi-parameter static fusion result. At the same time, the time series trend characteristics of the multi-dimensional data over the past 10-20 collection cycles (e.g., rate of temperature increase, voltage fluctuation amplitude, etc.) are calculated and assigned a trend weight coefficient of 0.2-0.3.

[0046] Finally, the static fusion results are weighted with a fusion weighting coefficient of 0.7-1.3, and then the weighted results of trend features are superimposed to form a high-dimensional data feature vector that can reflect the real-time operating status of the battery and include the parameter change trend. Finally, it is transmitted to the battery safety status assessment module for subsequent safety analysis.

[0047] The battery safety status assessment module receives data from the battery operation parameter sensing and fusion module, maps real-time parameters to corresponding safety status intervals through a preset multi-scale safety threshold matrix, generates battery safety status assessment results, and sends them to the risk warning decision module.

[0048] Specifically, the main function of the battery safety status assessment module is to receive fused data from the battery operating parameter sensing and fusion module, and using a preset multi-scale safety threshold matrix, accurately map real-time parameters to corresponding safety status intervals, thereby generating a comprehensive battery safety status assessment result, providing a direct basis for risk warning decisions. This module comprehensively considers multiple assessment parameters such as voltage, current, temperature, and pressure, assigning different weights to each parameter based on its impact on battery safety. This ensures that the assessment results accurately reflect the battery's safety status, avoiding the limitations of single-parameter assessments, effectively identifying potential battery safety risks, and laying the foundation for subsequent targeted protective measures.

[0049] During implementation, the first step is to determine the types and number of evaluation parameters, typically including four core parameters: voltage, current, temperature, and pressure. The weights of these parameters are set according to their degree of influence, with temperature having the highest weight (0.35-0.45), voltage at 0.25-0.35, current at 0.15-0.25, and pressure at 0.05-0.15. A pre-defined multi-scale safety threshold matrix sets multiple threshold levels for different parameters. For example, temperature thresholds are divided into three levels: normal (-20℃-45℃), warning (45℃-60℃), and dangerous (>60℃). Voltage thresholds are set according to battery type; for lithium-ion batteries, for instance, they are normal (3.0V-4.2V), warning (2.8V-3.0V or 4.2V-4.3V), and dangerous (<2.8V or >4.3V). After receiving the fused data, the module maps each real-time parameter to its corresponding safety threshold range. It then calculates the safety status score for each parameter using a mapping function. The score range is 0-100 points, with 80-100 points for the normal range, 40-80 points for the warning range, and 0-40 points for the danger range. Finally, the module calculates the battery safety status assessment value based on the weight of each parameter. An assessment value of 80-100 points indicates safety, 40-80 points indicates a warning, and 0-40 points indicates danger. The results are then sent to the risk warning decision module.

[0050] The risk warning and decision-making module generates corresponding risk warnings and control commands based on the battery safety status assessment results and a pre-built risk level-response strategy mapping table. These commands are then transmitted to the flame retardant network control module and the external linkage control module, respectively.

[0051] Specifically, the risk warning decision-making module, as the core of the battery safety protection system, primarily generates corresponding risk warning information and control commands based on the assessment results output by the battery safety status assessment module and a pre-built risk level-response strategy mapping table. This ensures that appropriate protective measures can be activated promptly when the battery experiences varying degrees of safety risks. This module also guarantees the accuracy and timeliness of the commands to prevent the escalation of safety incidents due to decision delays or incorrect commands. Simultaneously, it transmits the generated commands to the flame-retardant network control module and the external linkage control module, enabling multi-module collaborative response and improving overall protection efficiency.

[0052] In practical implementation, a pre-constructed risk level-response strategy mapping table divides risk levels into three levels: low, medium, and high, corresponding to battery safety status assessment values ​​of 80-100 points (low risk), 40-80 points (medium risk), and 0-40 points (high risk). The risk level weighting coefficients are set to 0.2-0.3 for low risk, 0.3-0.5 for medium risk, and 0.5-0.8 for high risk. Upon receiving a battery safety status assessment value, the module matches the assessment value with each risk level using a matching function. If the assessment value is 85 points, it matches the low-risk level, generating a low-risk warning message with a warning frequency set to once every 5-10 minutes, and generating preliminary control instructions, such as increasing parameter monitoring frequency. If the assessment value is 55 points, it matches the medium-risk level, generating a medium-risk warning message with a warning frequency set to once every 1-3 minutes, and generating enhanced control instructions, such as adjusting some parameters of the flame-retardant network. If the assessment value is 25 points, it matches the high-risk level, generating a high-risk warning message with a real-time alarm setting, and generating emergency control instructions, such as activating fire alarm linkage. Meanwhile, the risk instruction priority determination unit will determine the priority of the instructions. High-risk instructions are set to level 1, medium-risk instructions to level 2, and low-risk instructions to level 3. When multiple risks coexist, high-priority instructions are executed first to ensure that critical risks are dealt with in a timely manner.

[0053] The flame retardant network dynamic control module receives control commands from the risk warning decision module and, in conjunction with the optimization parameters output by the battery safety protection analysis module, dynamically adjusts the flame retardant release amount and flame retardant structural morphology of the multi-scale flame retardant network.

[0054] Specifically, the key function of the flame retardant network dynamic control module is to receive control commands from the risk warning decision module and, in conjunction with the optimization parameters output by the battery safety protection analysis module, dynamically adjust the flame retardant release amount and flame retardant structural morphology of the multi-scale flame retardant network to adapt to different battery safety risk conditions, ensuring that the flame retardant network always maintains optimal protective performance. The optimization parameters are derived by the battery safety protection analysis module through the integration of multi-source data and dynamic analysis calculations. The specific process is as follows: This module maintains bidirectional data interaction with the multi-scale flame retardant network construction module, continuously receiving its transmitted real-time operating data of the multi-scale flame retardant network, including the topological parameters, effective volume, flame suppression parameters, and overall protective performance of each level of flame retardant unit; simultaneously, it acquires the fused data feature vector (covering real-time and trend data of voltage, current, temperature, and pressure) output by the battery operating parameter perception fusion module, as well as the safety status assessment value and safety scores of each parameter generated by the battery safety status assessment module. Based on this multi-source data, the module combines the complexity of the battery pack's spatial layout, the protection requirements corresponding to the current risk level, and references historical fault data and real-time parameter trends. Through a pre-set algorithm, it comprehensively analyzes the current protective shortcomings of the flame-retardant network, the flame-retardant consumption rate, and structural adaptability. Finally, it calculates and generates optimized parameters, including the flame-retardant release rate adjustment threshold, optimized flame-retardant layer thickness / density values, and parameters for locally reinforced areas. Through precise control, this module can appropriately adjust the protection level when the battery presents a low risk, avoiding resource waste, and rapidly strengthen protection when a high risk occurs, effectively suppressing risk spread and ensuring the safety of the battery system.

[0055] During implementation, the module first receives control commands and optimization parameters, determines the types of adjustable parameters, mainly including four core parameters: flame retardant release rate, release amount, flame retardant layer thickness, and density. The weight coefficients of the adjusted parameters are set according to the risk level. For high risk, the weight coefficients of flame retardant release amount and flame retardant layer density are set to 0.4-0.5, and for low risk, they are set to 0.2-0.3. When a low-risk control instruction is received, the flame retardant release rate is adjusted to 0.5–1.0 g / min, the release amount is controlled within the range of 5–10 g, the flame retardant layer thickness is maintained at 2–3 mm, and the density is maintained at 1.2–1.5 g / cm³. When a medium-risk control instruction is received, the flame retardant release rate is increased to 1.0–1.5 g / min, the release amount is increased to 10–15 g, the flame retardant layer thickness is adjusted to 3–4 mm, and the density is increased to 1.5–1.8 g / cm³. When a high-risk control instruction is received, the flame retardant release rate is further increased to 1.5–2.0 g / min, the release amount is increased to 15–20 g, the flame retardant layer thickness is adjusted to 4–5 mm, and the density is increased to 1.8–2.0 g / cm³. Simultaneously, if a local safety hazard is detected in the battery, the flame-retardant network in the corresponding area will be strengthened. This involves increasing the local flame retardant release by 20%–30%, increasing the flame-retardant layer thickness by 1–2 mm, and coordinating adjustments to various parameters to ensure optimal protection of the entire flame-retardant network. This process of coordinating parameter adjustments to ensure optimal protection is achieved through multi-dimensional data interaction and dynamic adaptation, with the flame-retardant network collaborative optimization unit at its core. First, this unit collects real-time data on the local release increment from the precise flame retardant release adjustment unit, the regional thickness increase parameters from the dynamic flame-retardant structure adjustment unit, and the location and strengthening degree of the hazard area from the local strengthening unit. Simultaneously, it receives global protection effectiveness assessment data (calculated based on the multi-scale flame-retardant network comprehensive protection effectiveness formula) from the battery safety protection analysis module, clarifying the topological matching degree of each level of flame-retardant unit after local strengthening, the integrity of the effective volume coverage, and the synergy of flame suppression parameters. Based on this, a preset algorithm analyzes the potential chain reactions caused by adjustments to local parameters. For example, if the release of flame retardant in a core area increases, the release rate in adjacent non-core areas needs to be reduced accordingly to avoid excessive overall consumption. Simultaneously, the density parameters of the surrounding flame retardant layer are fine-tuned to maintain a balanced protection in three-dimensional space. If locally thickening the flame retardant layer may affect heat dissipation, the ventilation system parameters need to be checked for compatibility and fed back to the flame retardant structural morphology unit to optimize the porosity of the thickened area. Finally, the coordinated parameter instructions are synchronously sent to each functional unit to ensure that local reinforcement and global protection parameters complement each other, avoiding overlapping or weak areas and achieving dynamic optimization of the overall flame retardant network's protective effectiveness.

[0056] The collaborative external control module receives control commands from the risk warning and decision-making module and communicates with the fire protection system, ventilation system, and power management system to perform multi-system collaborative control.

[0057] Specifically, the collaborative external control module is primarily responsible for receiving control commands from the risk warning and decision-making module. It then communicates in real-time with the fire protection system, ventilation system, and power management system to achieve collaborative control across multiple systems, forming a comprehensive external safety protection system. This compensates for the shortcomings of the internal protection of the flame-retardant network and further enhances the safety protection capabilities of the battery system. By coordinating all external systems, this module ensures that when a safety risk arises with the battery, each system can simultaneously activate corresponding measures, avoiding ineffective protection due to asynchronous responses between systems.

[0058] In practice, after receiving control commands, the module communicates with various external systems according to the command level, with communication delays controlled within ±0.5 seconds. When a low-risk control command is received, a command is sent to the ventilation system to adjust the ventilation rate to 2-3 m / s to accelerate air circulation around the battery pack; a command is also sent to the power management system to closely monitor the charging current and voltage to maintain the current charging and discharging state. When a medium-risk control command is received, in addition to maintaining the ventilation system's ventilation rate at 2-3 m / s, a command is sent to the fire protection system to put it into standby mode, maintaining the fire sprinkler pressure at 0.8-1.0 MPa; a command is also sent to the power management system to reduce the charging current by 10%-20% to avoid excessive battery load. When a high-risk control command is received, a command is sent to the ventilation system to increase the ventilation rate to 3-4 m / s to accelerate the discharge of harmful gases; a command is sent to the fire protection system to activate the fire sprinkler system with a sprinkler flow rate set to 5-8 L / min; a command is sent to the power management system to immediately cut off the battery charging and discharging circuit to prevent excessive current from exacerbating the risk; and at the same time, continuous communication is maintained with each system to adjust the control command according to changes in risk.

[0059] The multi-scale flame retardant network construction module, battery operating parameter sensing and fusion module, battery safety status assessment module, risk early warning decision module, flame retardant network dynamic control module, and collaborative external control module are connected in sequence to form a complete battery safety protection system.

[0060] Preferably, the multi-scale flame retardant network construction module uses the following optimization model formula when constructing the multi-scale flame retardant network:

[0061]

[0062] in, This indicates the comprehensive protective effectiveness of a multi-scale flame-retardant network; This is the global adjustment coefficient, with a value ranging from 0.8 to 1.2. The number of layers in the flame-retardant network; For the first The weighting coefficient of each level is determined by the complexity of the battery pack's spatial layout; For the first The topological parameters of the tiered flame-retardant unit describe its spatial arrangement; For the first The effective working volume of the tiered flame-retardant unit; For the first The flame propagation suppression parameters of the tiered flame retardant unit.

[0063] Specifically, the material composition and structural parameters of the flame-retardant units in the multi-scale flame-retardant network construction module are clearly defined to improve the stability and specificity of flame-retardant performance. Global adjustment coefficient. The value ranges from 0.8 to 1.2, used to balance the overall effect of different levels of flame-retardant units. When the battery pack layout is complex, the value tends to be 1.2 to weaken the excessive influence of a single level; level weighting coefficient. The weighting is determined by the complexity of the battery pack's spatial layout. The core area, due to its higher protection priority, has a weighting coefficient set at 0.3-0.5, while the non-core area layers have a different weighting coefficient. This directly assigns differentiated influence weights to different levels. The topological parameters of each flame-retardant unit... It is a quantitative indicator describing its spatial arrangement characteristics, specifically in the form of a numerical expression based on spatial layout rules. For example, it assigns corresponding baseline values ​​according to structural types such as "closely arranged regular hexagons" and "nested cubic grids," and then combines them with the spacing between adjacent units (usually...). The parameters such as arrangement density are corrected to ultimately form a normalized value between 0.6 and 1.0 (the denser the arrangement and the more uniform the coverage, the closer the value is to 1.0). The reason why the topological parameters of the spatial arrangement can be multiplied with the weight, effective volume, and suppression parameters to obtain the protective effectiveness is that the parameters have achieved quantitative unification and physical meaning correlation: weight Reflecting the importance of hierarchy and the effective volume Reflects the actual protection coverage of the unit and flame suppression parameters. ; represents the flame retardancy of the material itself, while topological parameters This quantifies the "amplifying or weakening effect of structural arrangement on the protective effect." For example, for units of the same volume and flame retardant capability, a tightly nested arrangement ( (High value) can more effectively block the flame propagation path, and its actual protective effectiveness is far higher than that of a loosely arranged ( (Low value). The formula couples the four core elements of "hierarchical priority, structural rationality, coverage, and material performance" through multiplication operations, ultimately generating a quantitative index that comprehensively reflects the overall protective capability of the flame-retardant network. .

[0064] The flame-retardant unit uses composite flame-retardant materials, with inorganic flame-retardant components accounting for 40%-60%, organic flame-retardant components accounting for 30%-50%, and synergists accounting for 5%-10%. Inorganic flame-retardant components can be magnesium hydroxide, aluminum hydroxide, etc., while organic flame-retardant components can be phosphate esters, halogenated hydrocarbons, etc., and synergists can be antimony trioxide, zinc oxide, etc. This formulation allows the flame-retardant unit to achieve an oxygen index of 30%-35% and a vertical burning rating of V-0. Regarding structural parameters, the cross-sectional shape of the flame-retardant unit can be circular, square, or hexagonal. The cross-sectional dimensions are set according to the layer requirements. The side length or diameter of the cross-section of the core layer flame-retardant unit is 5-8 mm, while that of the non-core layer is 3-5 mm. The spacing between adjacent flame-retardant units is controlled at 10-15 mm to ensure that the flame propagation path is effectively blocked. During implementation, the flame-retardant components are first mixed according to the formula, and then melt-extruded and cooled to form flame-retardant units. Then, according to the preset layer layout and spacing, they are fixed inside the battery pack housing by adhesive or snap-fit. At the same time, the surface of the flame-retardant units is waterproofed with a waterproof coating thickness of 0.1-0.2mm to deal with possible electrolyte leakage from the battery pack and ensure the long-term effectiveness of the flame-retardant units.

[0065] Preferably, the battery operating parameter sensing and fusion module uses the following optimized model formula when performing feature fusion processing on the collected data:

[0066]

[0067] in, This represents the feature vector of the fused data; The fusion weighting coefficient has a value range between 0.7 and 1.3. Number of sensor types; For the first The weight of sensor data is determined by the importance of the sensor in battery safety monitoring; For the first A vector obtained after normalization of sensor data; This is the trend weighting coefficient; This is a time series trend feature vector for multi-dimensional data.

[0068] Specifically, the sensor performance and data processing algorithm of the battery operating parameter sensing and fusion module are optimized to improve the accuracy of data acquisition and the reliability of the fusion results. In the optimization model for feature fusion processing of the acquired data, the fusion weight coefficient γ is a normalized floating-point number between 0.7 and 1.3. Its value is dynamically adjusted according to the battery operating stage and safety monitoring requirements. When the battery is in a critical charging and discharging stage (such as the end of fast charging), more emphasis needs to be placed on the real-time status of parameters, and the value of γ is biased towards 1.3. When the operating state is stable, γ can be lowered to 0.7-0.9 to balance the influence of real-time data and trend features. This numerical form is determined through preliminary experimental calibration and adaptation to actual working conditions to ensure that the fusion weight adjustment of multi-source sensor data has both accuracy and flexibility. Time series trend feature vector of multi-dimensional data The following process is used to obtain the data: First, the original sequences of multi-dimensional data such as voltage, current, temperature, and pressure within the past 10-20 acquisition cycles are extracted. The sliding window method is used to calculate the rate of change (such as the rate of temperature increase and the amplitude of voltage fluctuation), stability index (such as the standard deviation of current), and abrupt change characteristics (such as the pressure surge node) of each dimension of data. Then, these trend indicators are normalized (mapped to the range of 0-1) and combined into a vector form in the order of "rate of change - stability - abrupt change characteristics" (for example, the temperature trend value is 0.8, the voltage trend value is 0.2, the current trend value is 0.5, and the pressure trend value is 0.1 in a certain cycle, which forms a vector of [0.8, 0.2, 0.5, 0.1]). Finally, a trend feature vector that can comprehensively reflect the time evolution law of multiple parameters is generated. The voltage sensor has a measurement range of 0-5V, an accuracy of ±0.01V, and a linearity error ≤0.1%; the current sensor has a measurement range of 0-100A, an accuracy of ±0.1A, and a response time ≤1ms; the temperature sensor has a measurement range of -40℃-125℃, an accuracy of ±0.5℃, and a resolution of 0.1℃; the pressure sensor has a measurement range of 0-1MPa, an accuracy of ±0.01MPa, and a repeatability error ≤0.2%. For data processing algorithms, a combination of a weighted average fusion algorithm and a Kalman filter algorithm is used. The weighted average fusion algorithm is used for preliminary fusion of multi-sensor data with the same parameter, while the Kalman filter algorithm is used to eliminate random noise and drift errors in the data. The filter coefficient is set to 0.05-0.1, and the state estimation error is controlled within 5%. During implementation, the sensor is first calibrated to ensure its performance parameters meet the requirements. Then, the sensor is connected to the data acquisition card, which has a 16-bit sampling bit and a transmission rate of 1Mbps. The sensor data is transmitted to the processor via the data acquisition card. The processor uses an ARM Cortex-A series or DSP series chip with an operating frequency of 800MHz-1.2GHz. The data is processed according to a preset algorithm. After processing, the generated data feature vector is stored in a local storage module with a capacity of 8-16GB and backed up to a cloud server for subsequent data traceability and analysis.

[0069] Preferably, the battery safety status assessment module uses the following optimized model formula when performing safety status assessment:

[0070]

[0071] in, This indicates the battery safety status assessment value; To evaluate the number of different types of parameters; For the first The weight of each evaluation parameter is determined by its degree of impact on battery safety; Indicates the first Real-time parameters Mapped to the corresponding security threshold range The mapping function outputs the corresponding security status score.

[0072] Specifically, the threshold update mechanism and evaluation model optimization of the battery safety status assessment module ensure that the evaluation results can adapt to parameter drift caused by battery aging and environmental changes. The multi-scale safety threshold matrix needs to be updated regularly, with the update cycle set according to the battery's usage time. The initial update cycle for a new battery is 3 months, followed by monthly updates. During each update, approximately 100-200 sets of normal battery operation data and 50-100 sets of fault data are collected. Statistical analysis is used to calculate the mean, standard deviation, and extreme values ​​of each parameter, redetermining the threshold range for each safety interval. The updated thresholds must undergo 10-20 verification tests, with a pass rate ≥95% before being put into use. Regarding evaluation model optimization, machine learning algorithms are introduced, employing support vector machines or neural network models. Historical fused data and corresponding safety status labels are used as training samples, with a sample size of no less than 1000 sets. The training iterations are set to 500-1000 times, with a learning rate of 0.01-0.05. After model training, the test set accuracy must reach over 90%. During implementation, the module initiates a threshold update process through a scheduled task, automatically extracting historical data from the storage module for analysis and calculation, generating a new threshold matrix and replacing the old matrix. At the same time, the evaluation model is retrained every quarter, adding new operational data and safety status data to the training samples, optimizing model parameters, improving the ability to identify new failure modes, and ensuring that the evaluation results are always accurate and reliable.

[0073] Preferably, the risk warning decision module uses the following optimized model formula when generating risk warning and control instructions:

[0074]

[0075] in, This represents the generated risk response instruction vector; Number of risk levels; For the first Weighting coefficients for each risk level; This indicates the battery safety status assessment value. With the Risk Level The matching function outputs the response strategy parameters corresponding to the risk level.

[0076] Specifically, the output methods of early warning information and the instruction execution feedback mechanism of the risk warning decision module are standardized to enhance the system's operability and closed-loop control capabilities. Early warning information output methods include local audible and visual alarms and remote terminal notifications. The sound intensity of the local audible and visual alarm is ≥85dB, and the light color is differentiated according to risk level: green for low risk, yellow for medium risk, and red for high risk. The flashing frequency is 1 time / second for low risk, 2 times / second for medium risk, and 4 times / second for high risk. Remote terminal notifications are implemented via SMS, APP push, and email. The notification content includes the risk level, occurrence time, involved battery unit number, and suggested handling measures, with a message transmission delay ≤10s. Regarding the instruction execution feedback mechanism, the module needs to receive execution status feedback information from the flame-retardant network control module and the external linkage control module. The feedback information includes instruction reception status, parameter adjustment progress, and final execution result. The feedback timeout is set to 30s. If no feedback is received within the timeout period, the module will resend the instruction, with a resend count ≤3 times. If no feedback is still received, a backup early warning mechanism is activated, increasing the early warning frequency and notifying maintenance personnel for on-site investigation. During implementation, first configure the parameters of the audible and visual alarm and the information of the receiving terminal for remote notification. Then, set the feedback information receiving port and timeout judgment logic in the module. When the control command is generated, the feedback timer is started synchronously, the feedback information is monitored in real time, and the command is judged as to whether it is executed successfully based on the feedback result. If the execution fails, the preset process is followed to retry or the backup mechanism is activated to ensure that the protection measures are effectively implemented.

[0077] Preferably, the flame-retardant network dynamic control module uses the following optimization model formula when dynamically adjusting the flame-retardant network:

[0078]

[0079] in, This represents the adjustment parameter vector of the flame-retardant network; The number of types of adjustable parameters; For the first The weighting coefficients of the adjustment parameters; Adjust Indicates according to control commands Optimization parameters output by the battery safety protection analysis module For the A function that adjusts various parameters.

[0080] Specifically, the interface protocols and linkage logic between the collaborative external control module and external systems are clearly defined to ensure the smoothness and consistency of multi-system collaborative response. The communication interfaces between the module and the fire protection system, ventilation system, and power management system adopt RS485 or Ethernet interfaces. The communication rate of the RS485 interface is 9600-115200bps, and the communication rate of the Ethernet interface is 100Mbps-1Gbps. The communication protocol adopts Modbus-RTU or TCP / IP protocol, and the data frame format is set according to the protocol standard, including start bit, data bit, parity bit, and stop bit, where the data bit is 8 bits, the parity bit is even parity, and the stop bit is 1 bit. In terms of linkage logic, when the module receives a high-risk command, it must first send a circuit cut-off command to the power management system, and after a delay of 0.5-1 second, send a sprinkler start command to the fire protection system, and simultaneously send an increase airflow command to the ventilation system to avoid exacerbating the risk of short circuits due to starting the sprinklers first. For medium-risk commands, it first sends a current reduction command to the power management system, and then sends a maintain airflow command to the ventilation system, while the fire protection system remains on standby. For low-risk commands, it only sends monitoring commands to the ventilation system and the power management system. During implementation, the communication interfaces of each external system are first configured to ensure that the interface protocol matches the module. Then, the correctness of the linkage logic is verified through simulation testing. The sending of commands at different risk levels is simulated, and the response sequence and time of each system are recorded. The response time deviation must be controlled within ±0.2 seconds. After the test is passed, the linkage logic parameters are written into the module's control program. At the same time, an interface failure emergency plan is established. When the communication of a certain system interface is interrupted, the module can automatically switch to the backup interface or adopt a hard-wired control method to ensure that the linkage function is not interrupted.

[0081] Preferably, the battery safety status assessment module includes: a voltage parameter mapping assessment unit, which receives battery voltage data transmitted by the battery operating parameter sensing and fusion module, and maps real-time voltage parameters to corresponding safety status intervals according to a preset voltage safety threshold matrix to generate a voltage safety status assessment result; a current parameter mapping assessment unit, which receives battery current data, maps real-time current parameters to corresponding safety status intervals according to a preset current safety threshold matrix, and outputs a current safety status assessment result; a temperature parameter mapping assessment unit, which receives battery temperature data, maps real-time temperature parameters to corresponding safety status intervals according to a preset temperature safety threshold matrix to obtain a temperature safety status assessment result; and a pressure parameter mapping assessment unit, which receives battery pressure data, maps real-time pressure parameters to corresponding safety status intervals according to a preset pressure safety threshold matrix to obtain a pressure safety status assessment result.

[0082] Specifically, the battery safety status assessment module consists of four parameter mapping assessment units. Each unit conducts a specific assessment of voltage, current, temperature, and pressure parameters to ensure accurate judgment of battery safety status from multiple dimensions. Each unit requires a preset safety threshold matrix with corresponding parameters. The voltage safety threshold matrix is ​​divided into ranges according to battery type. For example, the normal range for lithium-ion batteries is set to 3.0V-4.2V, the warning range is 2.8V-3.0V or 4.2V-4.3V, and the danger range is <2.8V or >4.3V. The current safety threshold matrix is ​​set according to the battery's rated current. Taking a battery with a rated current of 100A as an example, the normal range is 0A-100A, the warning range is 100A-120A, and the danger range is >120A. The temperature safety threshold matrix is ​​generally set as normal -20℃-45℃, warning 45℃-60℃, and danger >60℃. The pressure safety threshold matrix is ​​set according to the pressure bearing capacity of the battery casing. The normal range is 0MPa-0.3MPa, the warning range is 0.3MPa-0.5MPa, and the danger range is >0.5MPa. During implementation, each unit first receives the corresponding parameter data transmitted by the battery operation parameter sensing and fusion module, then calls the preset safety threshold matrix, and matches the real-time parameters with the threshold range through a mapping algorithm to generate a single-parameter safety status assessment result. The assessment result adopts a 0-100 point system, with 80-100 points for the normal range, 40-80 points for the warning range, and 0-40 points for the danger range. Finally, the assessment results of the four units are transmitted to the system core to provide sub-item basis for comprehensive safety status assessment.

[0083] Preferably, the risk warning decision module includes: a low-risk warning instruction generation unit, which generates corresponding low-risk warning information and preliminary control instructions based on a pre-built risk level-response strategy mapping table when the battery safety status assessment result is at a low-risk level; a medium-risk warning instruction generation unit, which generates medium-level risk warning information and enhanced control instructions based on a pre-built risk level-response strategy mapping table when the battery safety status assessment result is at a medium-risk level; a high-risk warning instruction generation unit, which generates high-risk warning information and emergency control instructions based on a pre-built risk level-response strategy mapping table when the battery safety status assessment result is at a high-risk level; and a risk instruction priority determination unit, which prioritizes the generated warning and control instructions of different risk levels to ensure that high-priority instructions are executed first when multiple risks coexist.

[0084] Specifically, the risk warning decision-making module comprises four functional units. Through hierarchical processing and priority determination, it ensures accurate and efficient output of risk response instructions. The low-risk warning instruction generation unit is activated when the battery safety status assessment value is between 80 and 100 points. It generates warning information and preliminary control instructions based on a mapping table. The warning information is sent every 5 to 10 minutes. The preliminary control instructions include increasing the parameter acquisition frequency to 20Hz and strengthening flame retardant network monitoring. The medium-risk warning instruction generation unit is activated when the assessment value is between 40 and 80 points. The generated warning information is sent every 1 to 3 minutes. The strengthened control instructions include adjusting the flame retardant release rate to 1.0 to 1.5 g / min and increasing the ventilation system speed to 2 to 3 m / s. The high-risk warning instruction generation unit is activated when the assessment value is between 0 and 40 points. The warning information is sent in real time. The emergency control instructions include increasing the flame retardant release rate to 1.5 to 2.0 g / min, activating the fire protection system, and cutting off the battery charging and discharging circuit. The risk instruction priority determination unit sets priority rules: high-risk instructions are priority level 1, medium-risk instructions are priority level 2, and low-risk instructions are priority level 3. When multiple instructions coexist, they are executed in priority order. If a high-risk instruction is executed, medium- and low-risk instructions are suspended until the high-risk instruction is resolved, at which point the corresponding level of instruction is resumed. During implementation, the module first receives the battery safety status assessment value, and activates the corresponding warning instruction generation unit according to the assessment value range. After each unit generates an instruction, it transmits it to the priority determination unit. The determination unit sorts the instructions according to the rules and sends the highest priority instruction to the flame-retardant network dynamic control module and the external linkage control module, respectively. At the same time, the instruction generation and sending time are recorded for subsequent traceability.

[0085] Preferably, the flame retardant network dynamic control module includes: a flame retardant release amount precision adjustment unit, which receives control commands from the risk warning decision module and, in conjunction with optimization parameters output by the battery safety protection analysis module, precisely adjusts the release rate and amount of flame retardant in the multi-scale flame retardant network; a flame retardant structure morphology dynamic adjustment unit, which dynamically changes the structure morphology of the flame retardant network according to control commands and optimization parameters, including adjusting the thickness and density parameters of the flame retardant layer; a flame retardant network local reinforcement unit, which, when a local safety hazard is detected in the battery, strengthens the flame retardant network in the corresponding area according to control commands to increase flame retardant protection capability; and a flame retardant network collaborative optimization unit, which coordinates parameter adjustments between different parts of the flame retardant network to ensure that the overall protective performance of the flame retardant network reaches its optimal level.

[0086] Specifically, the flame retardant network dynamic control module achieves precise and dynamic adjustment of the flame retardant network through the coordinated work of four functional units, ensuring optimal protective performance. The precise adjustment unit for flame retardant release rate, upon receiving control commands, determines the release rate and total amount based on optimized parameters: 0.5–1.0 g / min for low risk and 5–10 g for total amount; 1.0–1.5 g / min for medium risk and 10–15 g for total amount; and 1.5–2.0 g / min for high risk and 15–20 g for total amount. Through the cooperation of a solenoid valve and a flow sensor, the release rate is monitored and adjusted in real time. The flow sensor accuracy is ±0.5 g / min, ensuring the error is within the allowable range. The dynamic adjustment unit for flame retardant structure morphology adjusts the flame retardant layer parameters according to commands, with a thickness of 2–3 μm for low risk. The thickness of the flame retardant layer is 3-4 mm with a density of 1.5-1.8 g / cm³ for medium-risk areas and 4-5 mm with a density of 1.8-2.0 g / cm³ for high-risk areas. Parameter changes are achieved through a mechanical adjustment device. The flame retardant network local reinforcement unit detects local hazards using sensors and increases the release of flame retardant by 20%-30% and thickens the flame retardant layer by 1-2 mm in the affected areas. The flame retardant network collaborative optimization unit collects adjustment data from each unit and coordinates parameters through algorithms to avoid parameter conflicts. For example, when adjusting the thickness of the flame retardant layer, the release position of the flame retardant is simultaneously optimized to ensure comprehensive protection. During implementation, the module first receives control commands from the risk warning decision module and optimization parameters from the battery safety protection analysis module. Then, it distributes the commands and parameters to the corresponding functional units. After each unit performs the adjustment operation, the execution results are fed back to the collaborative optimization unit. The collaborative optimization unit verifies the overall protection effectiveness. If it does not meet expectations, the parameters are readjusted until the protection effectiveness meets the requirements of the current risk level.

[0087] like Figure 2 As shown, the battery safety protection system based on a multi-scale flame-retardant network operates by including the following steps:

[0088] Step S1: The battery operating parameter sensing and fusion module collects multi-dimensional operating parameters of battery voltage, current, temperature and pressure in real time through an integrated sensor array, and performs spatiotemporal alignment and feature fusion processing on the collected data.

[0089] Step S2: The processed data is transmitted to the battery safety status assessment module. The module maps the real-time parameters to the corresponding safety status interval based on the preset multi-scale safety threshold matrix, thereby generating the battery safety status assessment result.

[0090] Step S3: The risk warning decision module receives the battery safety status assessment results and generates corresponding risk warnings and control instructions through a pre-built risk level-response strategy mapping table.

[0091] Step S4: The flame retardant network dynamic control module receives the control command and, in conjunction with the optimization parameters output by the battery safety protection analysis module, dynamically adjusts the flame retardant release amount and flame retardant structural morphology of the multi-scale flame retardant network.

[0092] Step S5: The external control module receives control commands and communicates with the fire protection system, ventilation system, and power management system to carry out multi-system collaborative control.

[0093] Step S6: Each module continues to cycle through the above process, constantly monitoring and protecting the battery's safety status.

[0094] To address the lack of coordination in traditional protection systems, this system constructs a multi-scale flame-retardant network based on topology adaptive adjustment. Each level of flame-retardant unit is arranged in a three-dimensional nested configuration according to the spatial distribution characteristics of the battery pack, forming a three-dimensional protective structure. Compared to traditional single flame-retardant material protection, the protection dimension expands from a planar plane to three-dimensional space, significantly enhancing the protective effect. Simultaneously, it establishes communication with external control modules and systems such as fire protection, ventilation, and power management. When the risk warning decision module generates control commands, multiple systems can respond collaboratively. For example, when the battery temperature rises abnormally, the ventilation system activates to cool it down, the fire protection system prepares for emergency response, and all systems work closely together, changing the situation where traditional protection units operate independently and lack coordination.

[0095] To address the issues of inaccurate and dynamistic responses in traditional assessments, the system's battery operating parameter sensing and fusion module collects multi-dimensional data such as voltage, current, temperature, and pressure, and performs spatiotemporal alignment and feature fusion to provide a more comprehensive and accurate data foundation for subsequent assessments. The battery safety status assessment module utilizes a pre-set multi-scale safety threshold matrix to map real-time parameters to corresponding safety status ranges, breaking the limitations of traditional fixed threshold assessments and dynamically adjusting assessment standards according to the actual operating conditions of the battery. The risk warning and decision-making module generates precise control commands based on the assessment results through a pre-built risk level-response strategy mapping table, achieving graded responses from low to high risk. The flame retardant network dynamic control module can also adjust the flame retardant release amount and flame retardant structural morphology in real time according to commands and optimization parameters, making protective measures more aligned with actual risk conditions, effectively avoiding misjudgments, omissions, and untimely responses, comprehensively improving the battery's safety performance under complex operating conditions, and significantly reducing the risk of accidents.

[0096] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0097] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A battery safety protection system based on a multi-scale flame-retardant network, characterized in that, include: The multi-scale flame retardant network construction module is used to construct a multi-scale flame retardant network with a fractal structure based on the spatial distribution characteristics of the battery pack. Through the nested arrangement of flame retardant units at different levels in three-dimensional space, a three-dimensional flame retardant protection system is formed, and bidirectional data interaction is performed with the battery safety protection analysis module. The battery operating parameter sensing and fusion module collects multi-dimensional operating parameters of battery voltage, current, temperature and pressure in real time through an integrated sensor array, performs spatiotemporal alignment and feature fusion processing on the collected data, and transmits the processed data to the battery safety status assessment module. The battery safety status assessment module receives data from the battery operation parameter sensing and fusion module, maps real-time parameters to corresponding safety status intervals through a preset multi-scale safety threshold matrix, generates battery safety status assessment results, and sends them to the risk warning decision module. The risk warning and decision-making module generates corresponding risk warnings and control instructions based on the battery safety status assessment results and a pre-built risk level-response strategy mapping table. These instructions are then transmitted to the flame retardant network control module and the external linkage control module, respectively. The flame retardant network dynamic control module receives control commands from the risk warning decision module and, in conjunction with the optimization parameters output by the battery safety protection analysis module, dynamically adjusts the flame retardant release amount and flame retardant structural morphology of the multi-scale flame retardant network. The collaborative external control module receives control commands from the risk warning and decision-making module and communicates with the fire protection system, ventilation system, and power management system to perform multi-system collaborative control.

2. The battery safety protection system based on a multi-scale flame-retardant network according to claim 1, characterized in that, The multi-scale flame retardant network construction module uses the following optimization model formula when constructing the multi-scale flame retardant network: . in, This indicates the comprehensive protective effectiveness of a multi-scale flame-retardant network; This is a global adjustment coefficient, ranging from 0.8 to 1.2, used to balance the influence weight of each level of flame-retardant unit as a whole; The number of layers in the flame-retardant network; For the first The weighting coefficient of each level is determined by the complexity of the battery pack's spatial layout; For the first The topological parameters of the tiered flame-retardant unit describe its spatial arrangement; For the first The effective working volume of the tiered flame-retardant unit; For the first The flame propagation suppression parameters of the tiered flame retardant unit.

3. The battery safety protection system based on a multi-scale flame-retardant network according to claim 1, characterized in that, The battery operating parameter sensing and fusion module uses the following optimized model formula when performing feature fusion processing on the collected data: . in, This represents the feature vector of the fused data; The fusion weighting coefficient has a value range between 0.7 and 1.

3. Number of sensor types; For the first The weight of sensor data is determined by the importance of the sensor in battery safety monitoring; For the first A vector obtained after normalization of sensor data; This is the trend weighting coefficient; This is a time series trend feature vector for multi-dimensional data.

4. The battery safety protection system based on a multi-scale flame-retardant network according to claim 1, characterized in that, The battery safety status assessment module uses the following optimized model formula when performing safety status assessment: . in, This indicates the battery safety status assessment value; To evaluate the number of different types of parameters; For the first The weight of each evaluation parameter is determined by its degree of impact on battery safety; Indicates the first Real-time parameters Mapped to the corresponding security threshold range The mapping function outputs the corresponding security status score.

5. The battery safety protection system based on a multi-scale flame-retardant network according to claim 1, characterized in that, The risk warning and decision-making module uses the following optimized model formula when generating risk warning and control instructions: . in, This represents the generated risk response instruction vector; Number of risk levels; For the first Weighting coefficients for each risk level; This indicates the battery safety status assessment value. With the Risk Level The matching function outputs the response strategy parameters corresponding to the risk level.

6. The battery safety protection system based on a multi-scale flame-retardant network according to claim 1, characterized in that, The flame-retardant network dynamic control module uses the following optimization model formula when dynamically adjusting the flame-retardant network: . in, This represents the adjustment parameter vector of the flame-retardant network; The number of types of adjustable parameters; For the first The weighting coefficients of the adjustment parameters; Adjust Indicates according to control commands Optimization parameters output by the battery safety protection analysis module For the first A function that adjusts various parameters.

7. The battery safety protection system based on a multi-scale flame-retardant network according to claim 1, characterized in that, The battery safety status assessment module includes: The voltage parameter mapping and evaluation unit receives battery voltage data transmitted by the battery operating parameter sensing and fusion module, maps real-time voltage parameters to the corresponding safe state range according to the preset voltage safety threshold matrix, and generates voltage safety state evaluation results. The current parameter mapping and evaluation unit receives battery current data, maps real-time current parameters to the corresponding safe state range through a preset current safety threshold matrix, and outputs the current safety state evaluation result. Temperature parameter mapping and evaluation unit: This unit receives battery temperature data, maps real-time temperature parameters to the corresponding safe state range according to a preset temperature safety threshold matrix, and obtains temperature safety state evaluation results. The pressure parameter mapping and evaluation unit receives battery pressure data and maps real-time pressure parameters to the corresponding safe state range according to a preset pressure safety threshold matrix, thereby obtaining the pressure safety state evaluation result.

8. The battery safety protection system based on a multi-scale flame-retardant network according to claim 1, characterized in that, The risk warning decision module includes: The low-risk warning instruction generation unit is used to generate corresponding low-risk warning information and preliminary control instructions based on a pre-built risk level-response strategy mapping table when the battery safety status assessment result is at a low-risk level. The medium-risk warning instruction generation unit is used to generate medium-level risk warning information and enhanced control instructions based on a pre-built risk level-response strategy mapping table when the battery safety status assessment result is at a medium-risk level. A high-risk warning instruction generation unit is used to generate high-risk warning information and emergency control instructions according to a pre-built risk level-response strategy mapping table when the battery safety status assessment result is at a high-risk level. The risk instruction priority determination unit is used to determine the priority of early warning and control instructions of different risk levels, so as to ensure that when multiple risks coexist, the high-priority instructions are executed first.

9. The battery safety protection system based on a multi-scale flame-retardant network according to claim 1, characterized in that, The flame-retardant network dynamic control module includes: The flame retardant release amount precision adjustment unit is used to receive control commands from the risk warning decision module and, in combination with the optimization parameters output by the battery safety protection analysis module, precisely adjust the release rate and release amount of the flame retardant in the multi-scale flame retardant network. A flame-retardant structure dynamic adjustment unit is used to dynamically change the structure of the flame-retardant network according to control commands and optimization parameters, including adjusting the thickness and density parameters of the flame-retardant layer. The flame-retardant network local reinforcement unit is used to strengthen the flame-retardant network in the corresponding area according to the control command when a local safety hazard is detected in the battery, thereby increasing the flame-retardant protection capability. The flame retardant network collaborative optimization unit is used to coordinate the parameter adjustments among various parts of the flame retardant network to ensure that the protective performance of the entire flame retardant network reaches its optimal level.

Images