An emergency treatment system based on internal short circuit of solid-state battery

Abstract

The application relates to the technical field of solid-state battery safety protection, and discloses an emergency processing system based on internal short circuit of a solid-state battery. The system comprises a state monitoring module, a short circuit feature extraction module, a risk assessment module and an emergency strategy generation module. The state monitoring module collects voltage fluctuation data, temperature gradient distribution data and interface impedance change data of the solid-state battery in real time; the short circuit feature extraction module identifies voltage sudden drop rate, thermal runaway propagation gradient and ion migration abnormality coefficient through time domain and frequency domain analysis; the risk assessment module calculates a dynamic risk level; and the emergency strategy generation module generates a local current limiting, regional isolation or global power-off strategy according to the risk level. The system realizes early identification and hierarchical response of internal short circuit of the solid-state battery, and improves the safety and reliability of the battery system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of solid-state battery safety protection, in particular to an emergency processing system based on internal short circuit of solid-state battery. BACKGROUND

[0002] Solid-state batteries are considered as an important development direction of next-generation energy storage technology due to their high energy density and good safety. However, solid-state batteries still face the risk of internal short circuit in practical application. Such short circuit may be caused by multiple factors such as lithium dendrite growth, interface degradation, manufacturing defects or mechanical damage. Unlike traditional liquid electrolyte batteries, the internal short circuit of solid-state batteries develops more covertly and propagates faster. Once thermal runaway occurs, the consequences are more serious.

[0003] In the prior art, the monitoring of internal short circuit of the battery mainly depends on the threshold judgment of voltage and temperature parameters, such as setting upper limit of voltage or temperature for early warning or cutting off the circuit. This kind of method has obvious shortcomings in the face of complex internal state of solid-state battery: the single parameter response of voltage and temperature lags behind, it is difficult to capture the weak characteristics in the early stage of short circuit; the key parameters such as internal interface impedance change and ion migration behavior of solid-state battery are not effectively included in the monitoring system, so that the early failure cannot be identified; thirdly, the existing emergency strategy is mostly "one size fits all" power-off processing, which lacks a hierarchical response mechanism for different short circuit development stages, which may cause unnecessary system shutdown or insufficient response.

[0004] In addition, some studies attempt to use model prediction or artificial intelligence methods for battery fault diagnosis, but these methods usually rely on a large amount of historical data for training, have limited generalization ability, and have high computational complexity, making it difficult to realize real-time processing in embedded systems. The internal short circuit of solid-state battery is sudden and variable, which requires the processing system to have high real-time performance, high accuracy and high reliability. The existing technology has not effectively solved these problems. Therefore, it is urgent to develop a system that can monitor multi-dimensional operating parameters in real time, accurately identify short circuit characteristics, dynamically assess risk levels and generate corresponding emergency strategies to cope with the safety challenges brought by internal short circuit of solid-state battery. SUMMARY

[0005] The purpose of the present application is to provide an emergency processing system based on internal short circuit of solid-state battery to solve the problems raised in the background.

[0006] To achieve the above purpose, the present application provides an emergency processing system based on internal short circuit of solid-state battery, which comprises:

[0007] A state monitoring module for real-time acquisition of multi-dimensional operating parameters of solid-state battery, the multi-dimensional operating parameters including voltage fluctuation data, temperature gradient distribution data and interface impedance change data;

[0008] a short-circuit feature extraction module configured to perform time-domain and frequency-domain feature analysis on the multi-dimensional operating parameters to identify core feature indicators of internal short circuit of the solid-state battery, the core feature indicators including a voltage drop rate, a thermal runaway propagation gradient, and an ion migration abnormality coefficient;

[0009] a risk assessment module configured to calculate a dynamic risk level of the internal short circuit of the solid-state battery according to the core feature indicators, the dynamic risk level including a critical risk state, a diffusion risk state, and a failure risk state;

[0010] an emergency strategy generation module configured to generate a multi-level emergency response strategy based on the dynamic risk level, the multi-level emergency response strategy including a local current-limiting control strategy, a regional isolation control strategy, and a global power-off control strategy.

[0011] Preferably, the short-circuit feature extraction module is further configured to:

[0012] establish a mapping relationship between the core feature indicators and a material degradation degree of the solid-state battery;

[0013] determine a position coordinate and an influence range radius of the internal short circuit of the solid-state battery according to the mapping relationship;

[0014] transmit the position coordinate and the influence range radius to the risk assessment module.

[0015] Preferably, the risk assessment module is further configured to:

[0016] calculate a thermal runaway propagation rate according to the position coordinate and the influence range radius;

[0017] construct a three-dimensional risk evolution model in combination with the thermal runaway propagation rate and the core feature indicators;

[0018] predict a development trajectory of the internal short circuit of the solid-state battery through the three-dimensional risk evolution model.

[0019] Preferably, the emergency strategy generation module is further configured to:

[0020] determine an emergency response time window according to the three-dimensional risk evolution model;

[0021] optimize an execution timing sequence of the multi-level emergency response strategy based on the emergency response time window to generate an emergency control instruction sequence including the execution timing sequence.

[0022] Preferably, the system further comprises:

[0023] an actuator control module configured to parse the emergency control instruction sequence and drive multiple types of actuators, the multiple types of actuators including a variable impedance regulator, a phase change material injection device, and an electromagnetic shielding array;

[0024] a feedback adjustment module configured to monitor execution effects of the multiple types of actuators in real time and generate dynamic adjustment parameters.

[0025] Preferably, the actuator control module is further configured to:

[0026] determine a deployment position of the variable impedance regulator according to the position coordinates of the occurrence position and the influence range radius;

[0027] calculate an injection flow rate and an injection pressure of the phase change material injection device based on the dynamic risk level;

[0028] adjust a shielding strength gradient of the electromagnetic shielding array according to the thermal runaway propagation rate.

[0029] Preferably, the feedback adjustment module is further configured to:

[0030] collect real-time working parameters of the multiple types of actuators, the real-time working parameters including an impedance regulation accuracy, a material coverage rate, and a shielding efficiency;

[0031] compare deviations of the real-time working parameters from expected control targets and generate parameter correction instructions containing the deviations.

[0032] Preferably, the system further comprises:

[0033] a strategy optimization module configured to reconstruct the multi-level emergency response strategy according to the parameter correction instructions and the dynamic adjustment parameters;

[0034] the strategy optimization module is further configured to establish an emergency response effect evaluation index system, the emergency response effect evaluation index system including a risk suppression efficiency and an energy loss coefficient.

[0035] Preferably, the strategy optimization module is further configured to:

[0036] calculate a strategy optimization weight coefficient according to the risk suppression efficiency and the energy loss coefficient;

[0037] adjust parameter thresholds of the three-dimensional risk evolution model based on the strategy optimization weight coefficient;

[0038] update control parameter accuracy of the emergency control instruction sequence.

[0039] Preferably, the system further comprises:

[0040] A data storage module is configured to archive historical data of the multi-dimensional operation parameters, the core feature indicators, the dynamic risk levels and the multi-level emergency response strategies.

[0041] Compared with the prior art, the present application has the following advantages:

[0042] The present application improves the accuracy and early warning capability of internal short circuit identification in solid-state batteries through real-time acquisition and fusion analysis of multi-dimensional operation parameters. The system comprehensively utilizes voltage fluctuation, temperature gradient distribution and interface impedance change data, overcoming the limitations of traditional methods relying on a single parameter, and can more comprehensively reflect the internal state changes of the battery, avoiding false positives or false negatives.

[0043] The short circuit feature extraction module identifies core feature indicators such as voltage drop rate, thermal runaway propagation gradient and ion migration abnormal coefficient through joint analysis of time domain and frequency domain, which can effectively represent different stages and evolution trends of short circuit occurrence, providing reliable input for risk assessment.

[0044] The risk assessment module dynamically calculates the risk level according to the core feature indicators, distinguishing between critical risk, diffusion risk and failure risk, covering the whole process from potential failure to severe failure, so that the system can adapt to the safety requirements in different application scenarios.

[0045] The emergency strategy generation module starts the corresponding level of response strategy according to the dynamic risk level, including local current limiting, regional isolation and global power-off measures. This multi-level response mechanism avoids the problem of over-treatment or insufficient response, ensuring safety while maintaining system operation as much as possible, extending the available time of the device.

[0046] The entire system has high integration and real-time performance, is suitable for embedded platform deployment, can realize rapid decision-making without relying on a large amount of historical data, and reduces the dependence on computing resources. The system can be widely used in electric vehicles, energy storage power stations, portable electronic devices and other fields, enhancing the active safety protection capability of solid-state battery systems, prolonging the service life of batteries, reducing maintenance costs, and promoting the application of solid-state battery technology in high safety requirement scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 A timing diagram of the solid-state battery internal short circuit emergency handling system described in the present application;

[0048] Figure 2 A flowchart for the enhanced risk assessment module;

[0049] Figure 3 A solid-state battery emergency handling system operation result graph. DETAILED DESCRIPTION

[0050] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work are within the protection scope of the present application.

[0051] Please refer to Figure 1 The present application provides an emergency processing system based on internal short circuit of solid-state battery, which comprises:

[0052] The state monitoring module collects multi-dimensional operation parameters of the solid-state battery in real time through a multi-source sensor network. Voltage fluctuation data is obtained by a high-precision voltage sensor at a sampling frequency of 1000 times per second. Temperature gradient distribution data is monitored by a distributed thermocouple array to monitor temperature changes at different positions on the surface and inside the battery. Interface impedance change data is measured in real time by using an alternating current impedance spectrum technology to measure the impedance characteristics of the electrode and electrolyte interface. The short circuit feature extraction module performs time domain and frequency domain feature analysis on the collected multi-dimensional operation parameters. Time domain analysis uses wavelet transform to extract the mutation characteristics of the voltage signal. Frequency domain analysis identifies abnormal resonance frequencies in the impedance spectrum by using fast Fourier transform, so as to accurately identify three core characteristic indexes, namely, voltage drop rate, thermal runaway propagation gradient, and ion migration abnormality coefficient. The risk assessment module calculates the dynamic risk level according to the core characteristic indexes. When the voltage drop rate exceeds 50 mV / s, it is marked as a critical risk state. When the thermal runaway propagation gradient is greater than 10 ℃ / mm and the ion migration abnormality coefficient is less than 0.7, it is determined as a diffusion risk state. When all three indexes exceed the threshold value, it is classified as a failure risk state. The emergency strategy generation module generates a multi-level emergency response strategy based on the dynamic risk level. The critical risk state triggers a local current limiting control strategy to achieve preliminary suppression by reducing local current output. The diffusion risk state starts a regional isolation control strategy to block the heat propagation path by using a physical isolation circuit. The failure risk state executes a global power-off control strategy to cut off the energy supply of the entire battery system.

[0053] Embodiment 1: Please refer to Figure 2, receive multi-dimensional operating parameter data streams from the state monitoring module, which contain voltage fluctuation raw signals, temperature distribution matrices, and impedance spectrum information; the module uses a parallel processing architecture to synchronously analyze the multi-source data, voltage fluctuation data are processed through time domain differentiation to extract the change rate characteristics, temperature data are calculated through spatial gradient algorithm to obtain heat conduction characteristics, and impedance data are identified through frequency domain decomposition to identify the phase shift of characteristic frequency points. The mapping relationship between core feature indicators and the degradation degree of solid-state battery materials is established using machine learning methods. The neural network model built in the module takes the voltage sag rate, thermal runaway propagation gradient, and ion migration abnormality coefficient as input features, and calculates the output of the separator thickness decay rate and electrode active material loss rate through three hidden layers. The mapping model is trained using historical fault data and can inversely deduce the material state according to real-time operating characteristics.

[0054] Determine the occurrence position coordinates of the internal short circuit of the solid-state battery using multi-sensor fusion positioning technology. The voltage sensor arranged at different positions of the battery detects the signal propagation time difference, and the time difference of arrival algorithm is used to calculate the three-dimensional space coordinates of the abnormal point. At the same time, combined with the data of the temperature sensor array, the heat source positioning algorithm is used to verify the accuracy of the coordinates, and finally the position information containing the xyz coordinate system is output. Calculate the influence range radius by comprehensively analyzing the electro-thermal coupling. Take the occurrence position coordinates as the center of the circle, simulate the heat diffusion process through the heat conduction equation, and combine the specific heat capacity and thermal conductivity parameters of the battery material to calculate the thermal influence radius at different time points. At the same time, considering the propagation law of electrical characteristics, the influence range boundary is corrected according to the impedance change data.

[0055] Transfer the occurrence position coordinates and the influence range radius to the risk assessment module using a high-speed data bus protocol. The data packet contains coordinate data, radius value, time stamp, and data check code, and the transmission rate reaches 1 Gbps to ensure real-time performance. The risk assessment module starts to calculate the thermal runaway propagation rate immediately after receiving the data.

[0056] A three-dimensional risk evolution model is constructed by combining the thermal runaway propagation rate and core characteristic indicators. The model establishes a dynamic probability field with spatial coordinates, time dimension, and risk intensity as the axes. The thermal propagation rate is integrated as a time function, and the characteristic indicators are integrated as a spatial distribution function. The model simulates the dynamic changes of the risk field by solving partial differential equations. The risk value of each grid point is determined by the local temperature gradient, voltage change rate, and ion mobility. The development trajectory of internal short circuit in solid-state batteries is predicted by a three-dimensional risk evolution model using a probability prediction method. Monte Carlo simulation is used to simulate multiple possible development paths. Each simulation randomly generates different boundary condition combinations. The output results include the main direction trajectory of risk diffusion, the possible affected functional areas, and the probability of different risk levels appearing in the future time period. These prediction data provide decision-making basis for emergency strategy generation.

[0057] From feature extraction to risk prediction, the complete analysis chain, the short circuit feature extraction module continuously updates the mapping relationship database to improve the positioning accuracy, and the risk assessment module continuously optimizes the parameter settings of the three-dimensional model. The data interaction between the two modules uses a millisecond-level synchronization mechanism to ensure the real-time and accuracy of risk prediction, providing early warning capability ahead of fault development for the system. During the module processing process, a multi-level cache design is used. Raw data is first stored in the high-speed cache for preprocessing, feature extraction results are stored in the intermediate cache for model calling, and finally the prediction data is output to the shared memory for use by downstream modules. This architecture ensures the smoothness of the data processing process and avoids data blocking or loss phenomena. All computing processes are logged in detail, including feature extraction parameters, model calculation intermediate values, and prediction result data. These records provide important references for system performance optimization and subsequent upgrades. Through continuous operation and continuous learning, the system gradually improves the recognition accuracy of the internal short circuit features of solid-state batteries and the accuracy of risk prediction.

[0058] Example 2: Receive three-dimensional risk evolution model output data from the risk assessment module, which contains the probability distribution of the internal short circuit development trajectory of the solid-state battery and the threshold of key parameters; the time window calculation unit inside the module immediately starts the emergency response time window analysis, identifies the operable time interval from the current time to the risk irreversible point by comparing the risk diffusion curve with the system preset safety operation boundary, and finally calculates the best intervention period corresponding to each risk level. The execution timing of the multi-level emergency response strategy based on the emergency response time window needs to be calculated using a dynamic programming algorithm, which takes maximizing risk control effect and minimizing energy loss as dual objective functions, and takes local current limiting control strategy, regional isolation control strategy and global power-off control strategy as selectable decision variables; the algorithm traverses all possible strategy combinations and time arrangement sequences, evaluates the superposition effect of each sequence on the time axis, and finally outputs the starting time point, duration and switching condition of each strategy, forming a tightly linked protection chain in the time dimension.

[0059] Generating emergency control instruction sequences containing execution timing involves instruction encoding and protocol packaging process. The local current limiting control strategy is encoded as current adjustment instructions with time stamp, which contains target current value, current drop slope and duration parameters, which are dynamically generated according to risk level and current battery state; the regional isolation control strategy is converted into trigger instructions of physical isolation device, which specifies the number of isolators to be activated, trigger delay time and isolation maintenance time, ensuring the circuit connection of the specified area is cut off at the exact time; the global power-off control strategy generates multi-level power-off instruction sequence, which contains voltage grading drop curve parameters, final power-off state confirmation mechanism and system self-locking time setting.

[0060] All instructions are arranged into an ordered instruction stream according to the calculated execution timing, and each instruction is marked with high-precision time synchronization to ensure the coordinated operation between different execution mechanisms; the instruction sequence is transmitted to the execution mechanism control module through the safety communication protocol, and the transmission process adopts redundancy check and retransmission mechanism to ensure the integrity and reliability of the instructions. The emergency strategy generation module continuously obtains execution effect data from the feedback regulation module to optimize the time window calculation algorithm and strategy timing arrangement; the built-in strategy simulation unit in the module can simulate the execution effect before the actual issuance of instructions, further adjust the instruction parameters and time arrangement, and form a continuously self-optimizing emergency response mechanism.

[0061] Taking a practical operation scenario of a vehicle-mounted solid-state battery system as an example, the battery module suddenly detects an abnormal signal during vehicle driving; the state monitoring module detects that the voltage of No. 3 module fluctuates abnormally, the voltage value decreases from 4.2V to 3.8V within 0.1 seconds, and the infrared thermal imager shows that the surface temperature gradient of the module appears non-uniform distribution, with a maximum temperature difference of 15℃, and the electrochemical impedance spectrometer monitors that the electrode interface impedance at the characteristic frequency of 1000Hz increases significantly. The emergency strategy generation module receives the three-dimensional risk evolution model data transmitted by the risk assessment module, and the model prediction shows that the thermal runaway risk will spread from the current area to the adjacent module within 8 seconds; the module immediately starts the emergency response time window calculation, determines the best intervention time window as the 2nd to 6th second after detecting the abnormality by analyzing the intersection of the risk diffusion curve and the system safety boundary, and this time interval considers the mechanical response delay (1.2 seconds) and the control signal transmission time (0.3 seconds) of the actuator.

[0062] Based on the emergency response time window, the execution timing of the multi-level emergency response strategy is optimized, and the module uses a dynamic programming algorithm to calculate the strategy sequence: at t+2.0 seconds, start the local current limiting control strategy to limit the output current of No. 3 module from 200A to 50A; at t+3.5 seconds, start the regional isolation control strategy to activate the thermal isolation device and circuit breaker around No. 3 module; at t+5.8 seconds, start the global power-off control strategy to reduce the output voltage of the entire battery system in stages. This timing arrangement ensures seamless connection in time for each strategy, avoiding premature intervention affecting normal vehicle driving, and preventing late response leading to risk diffusion.

[0063] When generating the emergency control instruction sequence containing the execution timing, the module improves the time control precision to the microsecond level: the local current limiting instruction contains parameters such as target current value 50A, current drop slope 125A / s, and duration 2.8 seconds; the regional isolation instruction specifies the numbers of 3-1, 3-2, and 3-3 isolators that need to be activated, sets a trigger delay of 0.2 seconds and a minimum maintenance time of 4 seconds; the global power-off instruction contains curve parameters such as the voltage from 400V to 0V in stages, each stage down by 50V for 0.5 seconds of stability detection. All instructions are marked with nanosecond-level time synchronization, and the transmission reliability is ensured through redundant check coding. The instruction sequence is transmitted to the actuator control module through the CAN bus protocol, and the transmission process adopts a priority scheduling mechanism: the local current limiting instruction is transmitted in real time as the highest priority, the regional isolation instruction is the secondary priority, and the global power-off instruction is the basic guarantee level. The transmission data packet contains instruction type, execution timestamp, control parameters, and check code, and each data packet is 32 bytes in size, with a transmission rate of 1Mbps.

[0064] Throughout the entire emergency response process, the module continuously receives execution state feedback from the feedback adjustment module: when the local current limiting strategy is implemented, if the actual current drop curve deviates from the expected value by more than 5%, the module immediately generates correction instructions to adjust the subsequent strategy parameters; when the regional isolation strategy is executed, if the complete activation time of the isolation device is confirmed by the pressure sensor to be 0.1 seconds longer than expected, the module adjusts the starting time point of the global power-off strategy accordingly. The complete workflow of the emergency strategy generation module in a real scenario is shown: starting from receiving risk model data, through time window calculation, strategy timing optimization, instruction generation and transmission, an executable emergency control sequence is finally formed. The entire processing process is completed within milliseconds, reflecting the system's rapid response capability and fine control level in the face of sudden failures.

[0065] Referring to Figure 3 , the Figure 3 The complete operation process of the solid-state battery emergency handling system when detecting an internal short circuit anomaly is shown. Figure 3 The four subgraphs respectively show the changes of voltage, temperature, impedance, and the execution of emergency strategies. The upper left subgraph shows the change of battery voltage over time. Under normal conditions, the battery voltage remains stable at about 4.2V. When the system detects an anomaly (at about 2 seconds), the voltage drops rapidly from 4.2V to 3.8V within 0.1 seconds. This sharp voltage change is a typical feature of an internal short circuit. The system captures this change through a high-precision voltage sensor with a sampling frequency of 1000 times per second, providing key data for subsequent risk assessment. The upper right subgraph shows the change of battery temperature. Under normal conditions, the battery temperature fluctuates slightly at about 25°C. When an anomaly occurs, the temperature begins to rise significantly, with a maximum temperature difference of 15°C, indicating that a thermal runaway risk is forming. A distributed thermocouple array monitors temperature changes at different positions on the battery surface and inside, providing temperature gradient distribution data to the system. The lower left subgraph shows the change of electrode and electrolyte interface impedance. Under normal conditions, the impedance value fluctuates around 0.5Ω. When an anomaly occurs, the impedance at the characteristic frequency of 1000Hz increases significantly, indicating an abnormal ion migration process. The system uses AC impedance spectroscopy technology to measure interface impedance characteristics in real time, providing important parameters for short circuit feature extraction. The lower right subgraph shows the change of risk level and the execution time point of the emergency strategy. The system calculates the dynamic risk level based on three core feature indicators: voltage drop rate, thermal runaway propagation gradient, and ion migration anomaly coefficient. When the voltage drop rate exceeds 50mV / s, the system marks it as a critical risk state; when the thermal runaway propagation gradient is greater than 10°C / mm and the ion migration anomaly coefficient is less than 0.7, it is determined to be a diffusion risk state; when all three indicators exceed the threshold, it is classified as a failure risk state.

[0066] Embodiment 3: The mechanism of coordination between the mechanism control module and the feedback adjustment module, which receives the emergency control instruction sequence from the emergency strategy generation module, which contains multi-level control commands with precise time stamps and corresponding execution parameters; the instruction analysis unit inside the module uses a real-time operating system for instruction decoding, and realizes high-speed instruction stream processing through a field programmable gate array, converting abstract instruction parameters into specific drive signals, among which the control signal of the variable impedance regulator generates an adjustable voltage source output through a digital-to-analog converter, the control instruction of the phase change material injection device is converted into a pulse width modulation signal to drive the micro pump, and the adjustment command of the electromagnetic shielding array is generated through a radio frequency signal generator to generate an electromagnetic field of a specific frequency to control the waveform.

[0067] Driving multiple types of actuators requires differentiated driving strategies based on the physical characteristics of each actuator. The variable impedance regulator uses a closed-loop control method to dynamically adjust the impedance value by monitoring the current feedback in real time to match the target parameters, with a response time controlled at the microsecond level. The phase change material injection device uses multi-channel cooperative control to allocate the injection proportion of different nozzles according to the geometric characteristics of the target area, and adjusts the injection flow rate in real time through a pressure sensor. The electromagnetic shielding array uses beamforming technology to achieve directional shielding effect by adjusting the phase difference of multiple transmitting units, while monitoring the field intensity distribution to ensure shielding uniformity. The feedback adjustment module monitors the execution effect of multiple types of actuators in real time. This module collects the actual working state of the actuator through a high-precision sensor network. The actual impedance value of the variable impedance regulator is measured using a four-wire method to eliminate the influence of wiring resistance. The phase change material coverage state is detected by two-dimensional scanning with an infrared thermal imager and combined with an image processing algorithm to calculate the effective coverage rate. The electromagnetic shielding efficiency is measured in three-dimensional space using a near-field probe array and a field intensity decay distribution map is drawn. The real-time working parameters collected are compared with the expected control target, and the deviation value of each parameter is obtained using a difference calculation method, and processed according to priority.

[0068] The process of generating dynamic adjustment parameters uses an adaptive control algorithm to calculate the compensation amount based on the size and direction of the deviation. For impedance adjustment deviation, a proportional-integral regulator is used to generate a compensation signal, which is output to the impedance adjustment circuit through a digital-to-analog converter. For material coverage deviation, a fuzzy logic controller is used to adjust the injection parameters, and the volume of phase change material needed to be supplemented is calculated based on the area of the insufficient coverage area. For shielding efficiency deviation, a model predictive control algorithm is used to optimize the field intensity distribution, and the adjustment amount of shielding strength is calculated by the following formula:

[0069]

[0070] Where: represents the adjustment amount of shielding strength (unit is Tesla, T), It is the environmental medium influence coefficient. This represents the real-time monitored electric field strength (unit: volts per meter, or V / m). It is a variable (the unit is seconds, i.e., s), which is used in the integration process from arrive It changes continuously, used to represent every specific moment within the integration range. Let be the time decay function. and These represent the start and end times of the adjustment, respectively. It is the vacuum permeability (the unit is Tesla·meter / Ampere, i.e., T·m / A). Wave impedance (unit: ohm, Ω); based on the principle of electromagnetic induction, magnetic field shielding strength , It is the magnetic field strength, and the electric field strength With magnetic field strength satisfy This formula accumulates the historical impact of the electromagnetic field strength change rate through integral calculations, achieving dynamic adjustment of the shielding strength. The actuator control module determines the deployment location of the variable impedance regulator based on the location coordinates and the radius of influence. This process uses graph theory algorithms to analyze the topology of the battery system, calculates the optimal control path centered on the short-circuit point, and selects the node with the highest impedance adjustment sensitivity as the priority control target. Based on the dynamic risk level, the injection flow rate and injection pressure of the phase change material injection device are calculated, and a mapping function between the risk level and fluid parameters is established. The basic parameters are obtained by looking up a table, and then dynamic compensation is performed based on the real-time temperature to ensure that the phase change material achieves the best coverage effect in the designated area.

[0071] According to the thermal runaway propagation rate, the shielding strength gradient of the electromagnetic shielding array needs to be adjusted. A coupling model of electromagnetic field and thermal field is established to describe the interaction between thermal propagation and electromagnetic shielding, and the spatial distribution of shielding field strength is dynamically adjusted by real-time monitoring of temperature field changes. The adjustment of shielding strength uses gradient descent algorithm to gradually enhance the shielding strength based on the direction of thermal propagation, forming a suppression gradient in the opposite direction of thermal diffusion. The real-time working parameters of multiple types of actuators collected by the feedback adjustment module include impedance adjustment accuracy deviation value, material coverage rate difference percentage and shielding efficiency gap value. These parameters are sent to the comparator unit after digital filtering processing, and are compared with the preset tolerance range in real time to generate parameter correction instructions containing deviation direction and amplitude. The instruction transmission adopts the priority queue mechanism to ensure that critical deviations are processed in time. The feedback data is used by the feedback adjustment module to continuously optimize the monitoring algorithm and correction logic. The two modules exchange millisecond-level data through a high-speed data bus, and use the time-triggered protocol to ensure the real-time and determinacy of data transmission. All control instructions and feedback data are time-synchronized, ensuring the accuracy of the collaborative operation of the distributed system.

[0072] Taking an actual operation scenario of a solid-state battery pack for an electric vehicle as an example, the battery management system detects abnormal phenomena in the second module: the voltage sensor records that the voltage of the No. 4 battery cell drops from 3.65V to 3.25V within 0.2 seconds, the temperature monitoring system shows that the surface temperature gradient of the battery cell reaches 18℃ / cm, and the impedance monitoring unit detects an abnormal peak value of the interface impedance at a frequency of 2000Hz. The actuator control module receives the instruction sequence sent by the emergency strategy generation module, which contains three levels of control commands: first, the local current limiting instruction started 0.5 seconds after detecting the abnormality, which requires limiting the current of the branch where the No. 4 battery cell is located from 150A to 30A; second, the regional isolation instruction activated after 1.2 seconds, which specifies to enable the thermal isolation plate and circuit breaker around the No. 4 battery cell; and third, the global voltage reduction instruction executed after 2.8 seconds, which requires reducing the voltage of the entire battery pack from 600V to a safe voltage. The instruction decoding unit in the module immediately decodes and processes the received instruction stream, and uses the time slice round-robin scheduling algorithm of the real-time operating system to ensure that high-priority instructions are processed first: the local current limiting instruction is converted into an analog voltage signal output to the variable impedance regulator, generating a 0-5V control voltage corresponding to a 0-200Ω impedance adjustment range; the regional isolation instruction is compiled into a digital switch signal sent to the isolation device controller, specifying the specific relay number and action timing; and the global voltage reduction instruction is converted into a PWM waveform output to the main circuit regulator, setting the duty cycle to gradually reduce from 100% to 0%.

[0073] When the variable impedance regulator is driven to perform local current-limiting control, the module generates accurate control signals according to the instruction parameters: first, calculate the descending slope of the current from 150 A to the target current of 30 A, and set the gradual change rate to 240 A / s; then dynamically adjust the impedance value according to the real-time temperature data of the battery cell, and increase the adjustment amount by 0.5 Ω for every 1 °C increase in temperature; at the same time, monitor the voltage feedback during the adjustment process to ensure that the change in impedance does not cause a secondary voltage jump. When the phase change material injection device is controlled to perform regional isolation, the module determines the injection strategy according to the thermal distribution map: calculate the coverage area with a radius of 8 cm centered on the No. 4 battery cell, and allocate the injection proportion of the surrounding 6 nozzles; calculate the total amount of phase change material required according to the temperature gradient data, set the injection flow rate to 12 ml / s, and maintain the injection pressure at 0.3 MPa; adjust the injection angle in real time through infrared monitoring to ensure that the material covers the entire hot area.

[0074] When the electromagnetic shielding array is adjusted to perform protective measures, the module generates control parameters based on the thermal runaway propagation model: set a gradient distribution of the shielding field strength according to the direction of thermal propagation, set the shielding strength to 85 dB on the high-temperature side and gradually reduce it to 60 dB on the low-temperature side; dynamically adjust the emission frequency according to the real-time temperature change, gradually increase it from the basic 1.2 GHz to 2.4 GHz to enhance the shielding effect; at the same time, monitor the electromagnetic field distribution to ensure that it does not interfere with the normal work of other sensors of the battery management system. The feedback adjustment module synchronously monitors the actual operating state of each execution mechanism: measures the actual impedance value of the variable impedance regulator through a high-precision current sensor, compares it with the target value, and finds that the deviation is within ±0.8 Ω; uses a machine vision system to detect the actual coverage area of the phase change material, and finds that the coverage rate is 5.2% lower than the target value; measures the shielding efficiency through an electromagnetic field strength instrument, and finds that the actual value is 3.7% lower than the expected value. These monitoring data are uploaded to the control module in real time at a sampling frequency of 100 Hz.

[0075] When generating dynamic adjustment parameters, the module uses an adaptive algorithm to calculate the compensation amount: for impedance deviation, generate a compensation instruction that increases by 0.2 Ω every 0.1 second; for insufficient coverage, calculate the volume of phase change material that needs to be supplemented and adjust the injection parameters of the corresponding nozzles; for shielding efficiency gap, recalculate the field strength distribution map and optimize the working frequency of the emission unit. All adjustment parameters are time-stamped and priority-marked to ensure that critical parameters are adjusted first. During the entire response process, the module maintains a control cycle of milliseconds: collect the state of the execution mechanism every 5 milliseconds, generate an adjustment instruction every 10 milliseconds, and update the control parameters every 20 milliseconds. This fine control mechanism ensures that emergency measures can be accurately implemented and effectively suppress the spread of internal short circuit risks in solid-state batteries.

[0076] Example 4: Fine monitoring and dynamic adjustment of the working state of the actuator by the feedback adjustment module, which continuously collects real-time working parameters of the variable impedance regulator, phase change material injection device and electromagnetic shielding array through a multi-source sensing system; the impedance adjustment accuracy is measured by a high-precision bridge meter to obtain the actual impedance value at a sampling frequency of 2000 times per second, the material coverage is scanned by a multi-spectral imaging system to obtain the distribution state of the phase change material on the surface of the battery, and the shielding efficiency is measured by an electromagnetic field strength mapping system to measure the field strength attenuation ratio before and after shielding. The collected real-time working parameters are immediately sent to the data preprocessing unit for filtering and standardization processing, the impedance data are filtered by Kalman filter to eliminate measurement noise, the coverage image is extracted by edge detection algorithm to obtain the effective coverage area, and the electromagnetic field data are generated into a complete three-dimensional field strength distribution map by spatial interpolation method; the processed parameters are compared with the preset expected control target in real time, the impedance adjustment accuracy deviation is calculated by absolute difference, the material coverage deviation is analyzed by image pixel difference, and the shielding efficiency deviation is compared by field strength attenuation rate.

[0077] The process of generating parameter correction instructions adopts a multi-level decision mechanism. For impedance adjustment deviation, when the deviation value exceeds the allowed range, an impedance compensation instruction is generated, which includes compensation direction, compensation amount and compensation rate parameters; for material coverage deviation, a directional supplement instruction is generated according to the coordinates of the insufficient coverage area, which specifies the nozzle number and supplement flow rate that need to be strengthened; for shielding efficiency deviation, a field strength adjustment instruction is generated, which includes frequency modulation parameters and power adjustment value. Referring to Table 1, the monitoring and comparison of real-time working parameters of the actuator are shown.

[0078] Table 1: Monitoring table of real-time working parameters of the actuator

[0079]

[0080] The deviation degree is calculated by a normalization method, the impedance deviation degree is obtained by dividing the absolute difference between the actual value and the target value by the range, the material coverage deviation degree is calculated by area difference percentage, and the shielding efficiency deviation degree is determined by the ratio of the actual attenuation rate to the target attenuation rate; the calculated deviation degree data is compared with the preset tolerance threshold, and when any one of the deviation degrees exceeds the tolerance range, the correction instruction generation process is triggered immediately. The generated parameter correction instruction is packaged in a structured data format, each instruction contains a time stamp, a device number, a correction type, a correction parameter and an emergency level identifier; the instruction transmission adopts a priority queue mechanism, the correction instruction with high emergency level is sent first to ensure that the key parameter deviation is processed in time.

[0081] The feedback adjustment module has a built-in historical data learning function, continuously records the deviation trend and correction effect of each parameter, and optimizes the deviation degree calculation model and correction parameter generation rule through machine learning algorithm; the module regularly updates the tolerance range threshold, dynamically adjusts the allowed deviation range according to the aging degree of the actuator and the use environment. The whole implementation process forms a closed-loop control cycle of monitoring-comparison-correction, a two-way data channel is established between the feedback adjustment module and the actuator control module, and the monitoring data and correction instructions are transmitted in real time; the module adopts redundant design to ensure the reliability of monitoring data, and any sensor failure will immediately trigger the standby sensing unit to take over the monitoring task. The execution effect of the correction instruction is monitored and recorded in real time to form a correction effect evaluation report; these reports are used to optimize the generation accuracy of subsequent correction parameters, continuously improving the response accuracy and stability of the control system. The module also has a self-diagnosis function, which can identify systematic deviations caused by performance degradation of the actuator and provide early warning of equipment components that need maintenance or replacement.

[0082] Under complex working conditions, the feedback adjustment module can coordinate the collaborative correction of multiple actuators. When it is monitored that there is mutual influence between impedance adjustment, material coverage and shielding efficiency, joint correction instructions will be generated to ensure that the adjustment operations of each actuator cooperate with each other rather than interfere with each other; this collaborative control capability is realized through multivariate optimization algorithm, which considers the coupling relationship between parameters and the priority of control targets. The real-time data processing capability of the module supports millisecond-level response, the whole process time delay from parameter acquisition to instruction generation is controlled within 10 milliseconds, ensuring rapid response to sudden deviations; the data processing unit uses parallel computing architecture to process multiple sensing data simultaneously, ensuring the running stability of the system under high load working conditions.

[0083] Example 5: This example focuses on the continuous improvement function of the strategy optimization module to the emergency handling capability of the system, which receives parameter correction instructions and dynamic adjustment parameters from the feedback adjustment module, which contains deviation records and corresponding adjustment logs generated by actuators during actual operation; the module uses data fusion technology to integrate multi-source information into time series data set, and ensures that parameter correction instructions and dynamic adjustment parameters are completely synchronized on the time axis through time alignment algorithm. The process of reconstructing multi-level emergency response strategy adopts a strategy optimization algorithm based on reinforcement learning, which uses historical execution effect data as training set to establish the mapping relationship between strategy selection and execution result; the parameter adjustment of local current limiting control strategy is optimized based on current control accuracy deviation data, the trigger condition of regional isolation control strategy is corrected according to isolation effect feedback, and the hierarchical parameters of global power-off control strategy are recalibrated according to energy loss records. The algorithm continuously updates the node parameters in the strategy decision tree through iterative calculation, so that the emergency strategy is more suitable for actual operating conditions.

[0084] The establishment of the emergency response effect evaluation index system requires the definition of quantitative evaluation standards. The risk suppression efficiency is obtained by calculating the ratio of the reduction of risk diffusion area before and after the emergency intervention to the intervention time, which reflects the ability of the control system to suppress risk diffusion per unit time. The energy loss coefficient is quantified by measuring the percentage of the total energy dissipated during the emergency process to the total energy storage capacity of the system, which reflects the degree of influence of emergency measures on the energy integrity of the system. The two indicators together constitute the core dimension of the evaluation system and provide an objective measurement benchmark for strategy optimization. The calculation of the strategy optimization weight coefficient based on the risk suppression efficiency and energy loss coefficient uses a multi-objective optimization algorithm. The algorithm finds the optimal balance point between the two indicators through Pareto frontier analysis and dynamically allocates the weight proportion according to the safety requirements and economic requirements of the actual application scenario. In high-risk scenarios, the risk suppression efficiency is given a higher weight coefficient, while in normal operating conditions, more emphasis is placed on the optimization of the energy loss coefficient. The calculated weight coefficient serves as an important basis for strategy adjustment and guides the subsequent model parameter optimization process.

[0085] Adjusting the parameter thresholds of the three-dimensional risk evolution model based on the strategy optimization weight coefficient requires the establishment of a parameter sensitivity analysis model. First, identify the parameter set with the highest correlation to the risk suppression efficiency and energy loss coefficient in the model, including the thermal conductivity coefficient boundary value, voltage mutation threshold, and ion mobility critical value. Then, according to the weight coefficient, recalculate the allowed fluctuation range of these parameters. In high-risk weight configuration, the parameter thresholds are appropriately tightened, while in high-economic weight configuration, the parameter tolerance is relaxed. The adjusted parameter thresholds make the risk evolution model more consistent with the actual optimization goal. Updating the control parameter accuracy of the emergency control instruction sequence involves improving the instruction generation algorithm. The current control accuracy of the local current limiting instruction is improved from milliamperes to microamperes by increasing the resolution of the digital analog converter to achieve more precise current regulation. The time control accuracy is optimized from milliseconds to microseconds using high-precision clock synchronization technology to ensure the timing accuracy of instruction execution. The spatial positioning accuracy is improved from millimeters to microns by introducing laser ranging and visual positioning technology to improve the positioning accuracy of the actuator. These accuracy improvements make the emergency control more precise and efficient.

[0086] The strategy optimization module also establishes a long-term learning mechanism, continuously collects policy execution effect data and updates the optimization model, and maintains the adaptability of the strategy through periodic retraining; the built-in performance evaluation unit generates an optimization effect report regularly, records the changes in performance indicators before and after strategy optimization, and provides a reference basis for subsequent optimization direction. The entire optimization process forms a closed-loop learning system, allowing the emergency handling capability to be continuously improved as operational experience accumulates. The module uses a distributed computing architecture to process massive optimization data, uses parallel computing technology to accelerate the strategy optimization process, and ensures that the system can respond in real time to changes in operating conditions; all optimization operations are logged in detail, including numerical changes before and after parameter adjustment, optimization algorithm calculation process, and final effect evaluation data, which provide important references for system maintenance and subsequent upgrades. Through continuous strategy optimization, the system gradually establishes an emergency response knowledge base that adapts to different operating conditions, continuously improving the handling capability for internal short circuit faults in solid-state batteries.

[0087] In a specific battery system operation instance, the strategy optimization module found through analysis of historical processing data that when the risk suppression efficiency reached 85% or more, the energy loss coefficient often exceeded 12%. The module immediately adjusted the optimization weights, increasing the weight coefficient of risk suppression efficiency from 0.6 to 0.8 and correspondingly reducing the weight of the energy loss coefficient. This adjustment made the subsequent emergency strategy focus more on safety performance, appropriately relaxing the constraints on energy loss while ensuring risk control effectiveness. The module also updated the parameter thresholds of the three-dimensional risk evolution model, adjusting the alarm threshold of the thermal conductivity coefficient from 0.25 W / m·K to 0.22 W / m·K and the voltage jump threshold from 50 mV / s to 45 mV / s, making the model more sensitive to risk signals. These adjustments enable the system to detect potential risks earlier, providing more time for emergency handling.

[0088] The precision improvement of the control instruction sequence is specifically reflected in the following: the current adjustment step of the local current limiting instruction is reduced from 5 mA to 1 mA, the time synchronization precision is improved from ±1 ms to ±0.1 ms, and the positioning error of the actuator is reduced from ±2 mm to ±0.5 mm. These improvements make the emergency control action more precise and accurate, reducing additional energy loss caused by rough control. Through continuous optimization, the system gradually forms an adaptive emergency handling capability for different operating conditions, continuously improving operational efficiency while ensuring safety. The entire optimization process is completely driven by actual operational data, ensuring that improvement measures always meet actual needs and avoiding over-optimization or under-optimization.

[0089] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A system for emergency handling based on internal short circuit of solid-state battery, characterized in that, Comprise: A state monitoring module for real-time acquisition of multi-dimensional operating parameters of the solid-state battery, the multi-dimensional operating parameters including voltage fluctuation data, temperature gradient distribution data, and interface impedance change data; A short-circuit feature extraction module for time-domain and frequency-domain feature analysis of the multi-dimensional operating parameters to identify core feature indicators of internal short-circuit of the solid-state battery, the core feature indicators including voltage drop rate, thermal runaway propagation gradient, and ion migration abnormality coefficient; A risk assessment module for calculating a dynamic risk level of internal short-circuit of the solid-state battery based on the core feature indicators, the dynamic risk level including critical risk state, diffusion risk state, and failure risk state; An emergency strategy generation module for generating a multi-level emergency response strategy based on the dynamic risk level, the multi-level emergency response strategy including local current-limiting control strategy, regional isolation control strategy, and global power-off control strategy; The risk assessment module is further configured to: Calculate thermal runaway propagation rate based on occurrence location coordinates and influence range radius; Construct a three-dimensional risk evolution model combining the thermal runaway propagation rate and the core feature indicators; Predict the development trajectory of internal short-circuit of the solid-state battery through the three-dimensional risk evolution model.

2. The solid-state battery internal short-circuit emergency handling system according to claim 1, characterized by, The short-circuit feature extraction module is further configured to: Establish a mapping relationship between the core feature indicators and the material degradation degree of the solid-state battery; Determine occurrence location coordinates and influence range radius of internal short-circuit of the solid-state battery based on the mapping relationship; Transfer the occurrence location coordinates and influence range radius to the risk assessment module.

3. The solid-state battery internal short-circuit emergency handling system according to claim 1, characterized by, The emergency strategy generation module is further configured to: Determine an emergency response time window based on the three-dimensional risk evolution model; Optimize the execution timing of the multi-level emergency response strategy based on the emergency response time window to generate an emergency control instruction sequence containing execution timing.

4. The solid-state battery internal short-circuit emergency handling system according to claim 3, characterized by, Further comprise: An execution mechanism control module for parsing the emergency control instruction sequence and driving multiple types of execution mechanisms, the multiple types of execution mechanisms including variable impedance adjuster, phase change material injection device, and electromagnetic shielding array; A feedback adjustment module for real-time monitoring of the execution effect of the multiple types of execution mechanisms and generating dynamic adjustment parameters.

5. The solid-state battery internal short-circuit emergency handling system according to claim 4, characterized by, The execution mechanism control module is further configured to: Determine the deployment location of the variable impedance adjuster based on the occurrence location coordinates and influence range radius; Calculate the injection flow rate and injection pressure of the phase change material injection device based on the dynamic risk level; Adjust the shielding intensity gradient of the electromagnetic shielding array based on the thermal runaway propagation rate.

6. The solid-state battery internal short-circuit emergency handling system according to claim 5, wherein The feedback adjustment module is further configured to: Acquire real-time working parameters of the multiple types of execution mechanisms, the real-time working parameters including impedance adjustment accuracy, material coverage rate, and shielding efficiency; Compare the deviation degree of the real-time working parameters from the expected control target to generate parameter correction instructions containing deviation degree.

7. The solid-state battery internal short-circuit emergency handling system according to claim 6, characterized by, Further comprise: A strategy optimization module for reconstructing the multi-level emergency response strategy based on the parameter correction instructions and the dynamic adjustment parameters; The strategy optimization module is further configured to establish an emergency response effect evaluation index system, the emergency response effect evaluation index system including risk suppression efficiency and energy loss coefficient.

8. The solid-state battery internal short-circuit emergency handling system according to claim 7, characterized by, The strategy optimization module is further configured to: calculate a strategy optimization weight coefficient according to the risk suppression efficiency and the energy loss coefficient; adjust a parameter threshold of the three-dimensional risk evolution model based on the strategy optimization weight coefficient; update a control parameter precision of the emergency control instruction sequence.

9. The solid-state battery internal short-circuit emergency handling system of claim 8, wherein, Further comprising: a data storage module configured to archive historical data of the multi-dimensional operation parameters, the core feature indicators, the dynamic risk levels, and the multi-level emergency response strategies.

Images