An Optimization Method for Suppressing Temperature Rise in Supercapacitors

By collecting and analyzing the terminal voltage, current, and temperature of the supercapacitor, and combining the core structure parameters, the timing of the charging and discharging pulses and the intervention of heat dissipation are dynamically adjusted. This solves the problem of lagging identification of deep hot spots in supercapacitors and improves the stability of temperature rise control and transient power output capability.

CN123121928APending Publication Date: 2026-11-17FENGBANGBANG INTELLIGENT TECH (TIANJIN) CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202611081179.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies cannot effectively identify deep hot spots in the supercapacitor core, resulting in delayed hot spot suppression, which affects device lifespan. Furthermore, the passive derating strategy adopted for safety sacrifices transient power output capability.

Method used

By collecting the terminal voltage, bidirectional current, and shell temperature of the supercapacitor, and combining the core structure parameters and radial thermal resistance distribution, the internal thermal state distribution relationship is established, the thermal state is dynamically updated, and the timing of charging and discharging pulses and the sequence of external heat dissipation intervention are adjusted according to these states. With the help of graded power limiting control and protection mechanisms, the timely identification and intervention of deep hot spots can be achieved.

Benefits of technology

This technology enables the identification of deep hot spots in the core before the outer casing temperature rises significantly, avoiding hot spot suppression lag, reducing the impact on transient power output capability, and improving the temperature rise control stability and operational continuity of the supercapacitor.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123121928A_ABST
    Figure CN123121928A_ABST
Patent Text Reader

Abstract

This invention relates to the field of thermal management and charge / discharge control technology for supercapacitors, and discloses an optimization method for suppressing temperature rise in supercapacitors. This method addresses the problems of traditional methods, such as the difficulty in timely suppressing deep hot spots, and the excessive sacrifice of high transient power density by existing passive derating strategies. The method first collects and organizes the terminal voltage, bidirectional current, and shell temperature during the operation of the supercapacitor. It then establishes the internal thermal state distribution relationship by combining the core structure parameters, radial thermal resistance distribution, and shell boundary conditions. The method continuously updates the internal thermal state at different radial positions of the core and adjusts the charging / discharging pulse timing, polarity alternation sequence, and external heat dissipation intervention sequence based on the update results. When the thermal state exceeds the limit or the operation is abnormal, it performs graded power limiting control and disconnection protection, thereby balancing internal temperature rise suppression and transient power support capabilities.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of thermal management and charge / discharge control technology for supercapacitors, specifically to an optimized method for suppressing temperature rise in supercapacitors. Background Technology

[0002] In wind power generation, supercapacitors are often used for yaw, pitch drive, transient power support, and low voltage ride-through of wind turbines. This typically involves high-frequency, high-rate pulse charging and discharging, which can easily cause temperature rises during rapid switching. Because supercapacitors use a core made of alternating electrodes and separators, they have a large internal radial thermal resistance. Heat transfer from deep within the casing is delayed through the core, making it difficult for external temperature measurements to accurately reflect the true internal thermal state. Existing patent applications CN112926186B propose an online method for estimating the internal temperature of a single supercapacitor cell, using a shell temperature and electrothermal coupling model. CN111628546B also proposes a supercapacitor charging and discharging method and device, controlling the charging and discharging process based on state parameters.

[0003] While existing technologies have addressed the temperature rise problem of supercapacitors from the perspectives of temperature estimation and charge / discharge control, they still have the following shortcomings. Firstly, the core of a supercapacitor is often formed by winding electrodes and a separator, resulting in a large radial thermal resistance within the device. Heat generated deep within the core cannot be conducted outwards. Therefore, control methods based on shell temperature detection or conventional state parameters cannot accurately reflect the true internal thermal state, especially failing to detect deep-seated hot spots in a timely manner. This leads to situations where the temperature in localized areas is already high when external temperature rise is detected, increasing the risk of electrolyte thermal degradation. Untimely hot spot suppression also affects the device's lifespan. Secondly, to ensure operational safety, existing technologies typically employ early current limiting or derating for protection. While this reduces temperature rise, it impacts the transient power output capability of supercapacitors during wind power generation, hindering their high-power response. Therefore, effectively suppressing internal temperature rise and preventing runaway deep hot spots while mitigating the impact of conservative derating on transient power output capability is a technical problem that needs to be solved in this field. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an optimized method for suppressing temperature rise in supercapacitors, which solves the problems of traditional methods that struggle to suppress deep hot spots in a timely manner, and the excessive sacrifice of high transient power density by existing passive derating strategies.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for optimizing temperature rise suppression in supercapacitors includes:

[0007] S1. Collect the terminal voltage, bidirectional current and shell temperature of the supercapacitor during operation, and organize them according to a unified time reference;

[0008] S2. Based on the core structure parameters, radial thermal resistance distribution, and shell boundary conditions, establish the internal thermal state distribution relationship from the core center to the shell.

[0009] S3. Based on the sorted terminal voltage, bidirectional current and shell temperature, combined with the internal thermal state distribution relationship, continuously update the internal thermal state of the core at different radial positions.

[0010] S4. Based on the updated internal thermal state, determine the conduction period, interval period, polarity alternation sequence, and external heat dissipation intervention sequence of the charging and discharging pulses;

[0011] S5. Adjust the timing of the charging and discharging pulses according to the determined sequence, and perform graded power limiting control and disconnection protection when the internal thermal state exceeds the limit or the operating state is abnormal.

[0012] Furthermore, S1 includes:

[0013] The sampling points are the positive terminal, negative terminal, and shell surface temperature of the target monitoring unit, and the time is marked by a unified clock of the controller.

[0014] Using high-frequency sampling time slices of voltage and current as a unified time slice, the most recent valid temperature value is used to maintain the temperature value that has not been updated.

[0015] The sampled data is organized into standardized operational data frames that include timestamps, terminal voltages, bidirectional currents, casing temperatures, data validity flags, and sampling status.

[0016] Verify and mark anomalies such as over-range voltage at the terminal, abnormal rate of change of bidirectional current, and sudden temperature change of the casing.

[0017] Writes are made to a circular buffer using overwrite and read-write isolation methods.

[0018] Furthermore, S2 includes:

[0019] Based on the inner and outer dimensions of the core, core height, electrode thickness, diaphragm thickness, number of winding layers, and shell interface state of the target monitoring unit, the heat conduction path from the core center to the inner wall of the shell is discretized in layers to determine the total equivalent heat capacity and total equivalent radial thermal resistance, and the node positions, node thicknesses, node heat capacities, and inter-node thermal resistances of each layer are allocated.

[0020] The basic structure is mapped and the boundary state is corrected by combining the shell temperature, ambient temperature, heat dissipation medium state, cooling level and installation posture.

[0021] Generate a dataset that includes heat transfer boundary level, boundary correction status, and current update time.

[0022] Furthermore, the basic structure is mapped and the boundary state is corrected by combining the shell temperature, ambient temperature, heat dissipation medium state, cooling level, and installation posture, including:

[0023] Based on the natural air cooling boundary, forced air cooling boundary, or liquid cooling boundary corresponding to the operating environment, and combined with the shell temperature, ambient temperature, operating status of the heat dissipation medium, cooling level, and device installation direction, the heat exchange boundary level and installation posture correction status are adjusted.

[0024] When the casing temperature remains higher than the ambient temperature, adjust the boundary heat dissipation capacity to the limited level;

[0025] When external cooling is activated and the casing temperature drops within a preset time window, the boundary heat dissipation capability is adjusted to the enhanced level.

[0026] Furthermore, S3 includes:

[0027] The running data window and the previous state of each radial level are retrieved according to a fixed update time;

[0028] Heat load is determined based on the effective time ratio of bidirectional current and the amplitude of terminal voltage fluctuation.

[0029] The heat input weights and heat transfer weights of the central layer, inner transition layer, outer transition layer, and boundary layer are adjusted in a layered manner;

[0030] Combine the stratified temperature change rate threshold, state acceptance rules, invalid data conservative update rules, temperature probe failure handling rules, and high and low temperature anomaly handling rules;

[0031] Update the current temperature value, heat accumulation level and outward heat transfer status level of each layer, and generate a dynamic internal thermal state sequence that includes the risk indicator of the central layer, the precursor indicator of thermal runaway and the global thermal state level.

[0032] Furthermore, heat load is determined based on the effective time ratio of bidirectional current and the amplitude of terminal voltage fluctuation, including:

[0033] Within the preset heat load discrimination window, the effective time percentage of bidirectional current exceeding the rated continuous current discrimination threshold is statistically analyzed, and combined with the fluctuation range of terminal voltage relative to rated voltage, the heat input status of the current pulse group is classified and judged.

[0034] When the effective time percentage reaches the medium sustained heat input criterion, the heat accumulation level of the central layer is increased;

[0035] When the effective time percentage reaches the strong continuous heat input criterion and the terminal voltage fluctuation enters the high pulse range, the heat accumulation level of the central layer and the inner transition layer is increased.

[0036] Furthermore, S4 includes:

[0037] Based on the temperature status of the central layer, the heat accumulation status of the inner transition layer, the temperature rise status of the boundary layer, the global thermal status level, the risk indicators of the central layer, and the signs of thermal runaway, a state machine is constructed that includes the normal operation status, the pre-regulation status, the enhanced heat dissipation status, and the protection switching preparation status.

[0038] The pulse conduction period, interval period, polarity arrangement sequence, and external heat dissipation intervention level are adjusted in stages.

[0039] Based on the state entry conditions, external heat dissipation level switching conditions, protection preconditions, state rollback conditions, computing power limited branches, and field missing branches, a pulse control sequence table containing the current running state and rollback lock flags is generated.

[0040] Furthermore, the pulse conduction period, interval period, polarity arrangement sequence, and external heat dissipation intervention level are adjusted in stages, including:

[0041] Based on the current operating status, the conduction period of a single high-heat input pulse and the interval between adjacent pulses are adjusted in stages.

[0042] When there is room for bidirectional power scheduling and the central layer continues to heat up and the heat accumulation level of the inner transition layer increases, adjust the order of charging pulses and discharging pulses.

[0043] When conditions for polarity alternation are not met, extend the interval period;

[0044] Based on the temperature rise rate of the central layer, the temperature rise status of the boundary layer, the temperature range of the central layer, and the risk indicators, the external heat dissipation intervention level is switched between standby, pre-start, and execution levels.

[0045] Furthermore, S5 includes:

[0046] Based on the current operating status and thermal risk level, a three-level power limit constraint is applied to the pulse release boundary, adjusting the cumulative conduction time, single conduction holding time, and average duty cycle respectively;

[0047] When the temperature exceeds the limit, the terminal voltage drops abnormally, or the operation is abnormal, switch to the disconnect protection state or the lock protection state.

[0048] Based on automatic recovery conditions or manual reset conditions, the status is rolled back, and combined with feedback adjustment rules, processing delay constraints and protection path priorities, execution status records and running status feedback information are generated.

[0049] Furthermore, in the event of temperature exceeding limits, abnormal voltage drop, or operational abnormality, the system switches to either a disconnected protection state or a locked protection state, including:

[0050] When the current temperature of the central layer reaches the disconnection threshold and the temperature change rate remains positive, it enters the disconnection protection state.

[0051] When the terminal voltage drops abnormally within a continuous control cycle and is accompanied by the activation of the central layer risk flag, it enters the disconnection protection state.

[0052] When the abnormal operation status flag indicates severe sampling distortion, control command mismatch, or external heat dissipation failure, and the global thermal status level is in protection mode, it enters the disconnect protection state.

[0053] If abnormal terminal voltage or continued rise in center layer temperature persists after the protection is disconnected, the system will switch to locked protection mode.

[0054] Compared with the prior art, the present invention provides an optimized method for suppressing temperature rise in supercapacitors, which has the following beneficial effects:

[0055] 1. This invention collects and organizes the terminal voltage, bidirectional current, and shell temperature during the operation of a supercapacitor. It then establishes the internal thermal state distribution relationship from the core center to the shell by combining core structure parameters, radial thermal resistance distribution, and shell boundary conditions. Based on the dynamically updated deep thermal state, it coordinates the conduction period, interval, polarity alternation sequence of the charge / discharge pulses, and the sequence of external heat dissipation intervention. Simultaneously, it incorporates graded power limiting control, disconnection protection, lockout protection, and recovery / retreat mechanisms. This allows for the identification of deep hotspot formation trends in the core before the shell temperature rise becomes significant, and preemptive intervention in deep heat accumulation. This avoids the lag in hotspot suppression caused by relying solely on shell temperature detection, while mitigating the excessive impact of early current limiting or uniform derating on transient power output capability. Thus, it achieves a balance between suppressing internal temperature rise and supporting transient power in the supercapacitor.

[0056] 2. This invention, by hierarchically modeling the heat conduction path from the core center to the outer shell, introduces state inheritance rules, abnormal conservative update rules, boundary state correction rules, and execution feedback adjustment mechanisms into the internal thermal state update. This enables continuous thermal state judgment and control actions when sampling anomalies, boundary heat transfer changes, and operating state fluctuations occur. This reduces the stability of the supercapacitor temperature rise control process, enhances the continuity of state switching, and improves the continuous controllability of the operating process caused by factors such as frequent pulse regulation switching, false triggering of heat dissipation intervention, and misjudgment of protection state due to single fluctuations, local measurement distortion, or short-term environmental changes. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of the process for optimizing the temperature rise suppression method of a supercapacitor according to the present invention;

[0058] Figure 2 This is a schematic diagram of the target monitoring unit's data acquisition and processing in this invention;

[0059] Figure 3 This is a radial thermal state layering model diagram of the present invention from the center of the core to the outer shell.

[0060] Figure 4 This is the logic diagram for the dynamic internal thermal state update of this invention;

[0061] Figure 5 This is a diagram showing the linkage between the pulse control state machine and external heat dissipation in this invention.

[0062] Figure 6 This is a flowchart of the graded power limiting control and protection recovery process of the present invention. Detailed Implementation

[0063] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] Example 1: Figures 1-6 An optimization method for suppressing temperature rise in supercapacitors is presented, including:

[0065] S1. Collect the terminal voltage, bidirectional current and shell temperature of the supercapacitor during operation, and organize them according to a unified time reference;

[0066] S2. Based on the core structure parameters, radial thermal resistance distribution, and shell boundary conditions, establish the internal thermal state distribution relationship from the core center to the shell.

[0067] S3. Based on the sorted terminal voltage, bidirectional current and shell temperature, combined with the internal thermal state distribution relationship, continuously update the internal thermal state of the core at different radial positions.

[0068] S4. Based on the updated internal thermal state, determine the conduction period, interval period, polarity alternation sequence, and external heat dissipation intervention sequence of the charging and discharging pulses;

[0069] S5. Adjust the timing of the charging and discharging pulses according to the determined sequence, and perform graded power limiting control and disconnection protection when the internal thermal state exceeds the limit or the operating state is abnormal.

[0070] Specifically, such as Figure 2 As shown: The target monitoring unit in a single supercapacitor cell or series-parallel module is selected as the data acquisition object. The target monitoring unit corresponds to the positive terminal, negative terminal, and shell surface temperature sampling location. The terminal voltage is defined as the instantaneous potential difference between the positive and negative terminals, in volts. The bidirectional current is defined as the instantaneous current flowing through the target monitoring unit, in amperes, where the charging direction is recorded as positive and the discharging direction as negative. The shell temperature is defined as the real-time temperature of the outer surface of the metal shell of the target monitoring unit, in degrees Celsius. The terminal voltage is obtained through an isolated voltage sampling circuit with a range of 0V to 3.2V. This range is used to cover the voltage range of a supercapacitor with a rated voltage of 2.7V within the normal operation range, short-time pulse fluctuation range, and sampling front-end input redundancy range. To distinguish between the measurement range and operation identification... The effective operating identification upper limit can be selected from 2.85V to 2.95V, for example, 2.9V. This value comes from the range of 1.05 to 1.09 times the rated voltage, which can be used as the short-term operating identification boundary. At the same time, it can avoid mistakenly taking the upper limit of the sampling range as the long-term allowable operating limit of the device. The bidirectional current is obtained through a Hall current sensor or a shunt in conjunction with an isolation amplifier circuit. The sampling range is from -1500A to +1500A. This range comes from the peak current coverage requirements of wind power transient power support and high-rate pulse charge and discharge conditions, while also ensuring that the sampling link does not saturate and distort. The shell temperature is obtained through a thermocouple, thermistor, or digital temperature probe. The sampling range is from -40℃ to 110℃. This range comes from the low-temperature start-up environment, normal operating environment, and high-temperature abnormal warning range of industrial equipment.

[0071] During the data acquisition process, the controller's internal unified clock is used as the main time reference, and the terminal voltage, bidirectional current, and casing temperature are time-stamped separately. The timestamp can be represented by a 64-bit unsigned integer, with the unit being microseconds. Because the terminal voltage and bidirectional current change rapidly under continuous high-frequency, high-rate pulse conditions, to cover the pulse conduction edge, sustaining interval, and polarity switching interval, the sampling frequency for the terminal voltage and bidirectional current can be selected from 8kHz to 15kHz, for example, 10kHz to 12kHz. This range is derived from the fact that it can reflect millisecond-level pulse changes within a sampling period of 125μs to 66.7μs, and is compatible with commonly used high-speed analog-to-digital converters. The sampling capabilities are matched; the shell temperature is affected by the shell's heat capacity and external heat exchange conditions, and its rate of change is significantly lower than that of electrical parameters. Therefore, the shell temperature sampling frequency can be selected from 5Hz to 20Hz, for example, 10Hz. This range is derived from the fact that the shell surface temperature usually does not undergo drastic changes that affect the judgment of thermal boundaries within a time scale of 0.05s to 0.2s. Using this sampling frequency can meet the boundary temperature tracking requirements and reduce the processing resource consumption. In order to unify the data under different sampling frequencies into the same processing link, the high-frequency sampling time slice of terminal voltage and bidirectional current is used as the unified time slice, and the corresponding data records are organized in each unified time slice.

[0072] The original sampling results are organized into standardized operational data frames according to a unified format. Each standardized operational data frame includes at least the following fields: timestamp, terminal voltage, bidirectional current, casing temperature, data validity flag, and sampling status. The terminal voltage and bidirectional current fields can be represented by 32-bit floating-point numbers, while the casing temperature field can be represented by 16-bit or 32-bit numerical values. The data validity flag indicates whether the current data frame meets the conditions for subsequent thermal status updates, and the sampling status field records out-of-range, signal interruption, synchronization anomaly, and interference flags. Since terminal voltage and bidirectional current are high-frequency samples, while casing temperature is a low-frequency sample, if no new casing temperature value is refreshed within a certain time slice, the most recent valid casing temperature value is used and maintained until the next casing temperature sampling update. To prevent old values ​​from being indefinitely extended when the temperature probe malfunctions, the upper limit of the continuous holding time can be selected from 0.5s to 2s, for example, 1s. This range is derived from the casing temperature sampling period and the allowable lag time of the boundary temperature. When the continuous holding time exceeds this upper limit, a temperature sampling failure status is written into the corresponding data frame.

[0073] During data processing, the collected results are front-end verified to distinguish between actual operating condition changes and measurement artifacts. For terminal voltage, if the detected value is below 0V or above the upper limit of the sampling circuit range, the current data frame is determined to be an abnormal frame and a resampling is triggered. The number of resampling times can be selected as 1 because in high-frequency sampling scenarios, it is necessary to prioritize compressing the confirmation time to avoid multiple resampling occupying the subsequent processing window. If the range is still exceeded for 3 consecutive frames, the abnormal state is retained and a conservative processing flag is written. The number of consecutive abnormal frames is set to 3 frames because when the sampling frequency is 10kHz, 3 frames correspond to an approximately 300μs confirmation window, which can distinguish between instantaneous glitches and persistent anomalies.

[0074] For bidirectional current, if the rate of change of current in two adjacent time slices exceeds a preset abnormal rate of change threshold, it is preferentially identified as sampling spike interference. The abnormal rate of change threshold can be selected from 40A / μs to 80A / μs, for example, 60A / μs. This range is determined based on the bus voltage level, the equivalent parasitic inductance of the circuit, the sampling bandwidth, and the maximum reasonable current ramp-up range under the pulse condition of the target device. When the bus voltage is between 600V and 1500V and the equivalent parasitic inductance of the circuit is between 10μH and 40μH, the reasonable rate of change of the real current in a single sampling period is usually lower than the above threshold. Therefore, exceeding the threshold can be used as a criterion for spike interference. When this anomaly occurs, adjacent historical valid data is called for slope smoothing correction. If more than 3 abnormal rate of change data occur consecutively in the same short time window, the corresponding time period is marked as an abnormal sampling segment.

[0075] For the shell temperature, if the temperature change rate of adjacent sampling points exceeds 10℃ / s to 20℃ / s, for example, 15℃ / s, it is preferentially determined to be an abnormal probe contact or external interference. This range is determined based on the surface temperature change characteristics of aluminum alloy or steel shells under common thickness and normal heat exchange conditions. Since the shell surface temperature usually does not change abruptly at this rate during normal thermal evolution, exceeding this range generally does not belong to a normal temperature rise process. When this abnormality occurs, the previous valid temperature value is maintained and recorded as a temperature abnormality status.

[0076] The standardized running data frames, after verification, are written into a circular buffer to form a continuously callable running data sequence. The length of the circular buffer can be selected from 4096 frames to 16384 frames, for example, 8192 frames. This range is determined comprehensively based on the sampling frequency, the short-time history window length, and the controller's storage capacity. When the sampling frequency of the terminal voltage and bidirectional current is 10kHz, 8192 frames correspond to approximately 819.2ms of continuous running history, which can cover the continuous pulse change process and meet the requirements of subsequent continuous thermal state updates for calling short-time historical data. Each frame in the buffer includes at least a frame sequence number and a frame checksum to prevent read / write misalignment and frame loss. The circular buffer adopts a first-in-first-out overwrite method, where newly arrived data frames overwrite the earliest data frames. At the same time, a read / write isolation method is used to avoid data conflicts between the sampling and processing process and the subsequent thermal state update process.

[0077] After the above processing is completed, a standardized operating data sequence is formed. Each data frame in this data sequence has a unique time identifier, a unified sampling benchmark, a complete field structure, and a traceable status record. It can be directly used as the external boundary condition and operating excitation condition when establishing the internal thermal state distribution relationship from the core center to the outer shell. To meet the real-time control requirements, the maximum processing delay from the completion of the original sampling to the writing of the standardized operating data frame into the ring buffer can be selected to be no more than 500μs, for example, no more than 200μs. This time range is determined according to the high-frequency sampling frequency and the timing allocation of the subsequent processing links. When the high-frequency sampling frequency is 10kHz to 12kHz, the sampling and processing process should be completed within 2 to 5 consecutive sampling cycles, thereby reserving processing time for subsequent internal thermal state updates and pulse timing adjustments.

[0078] Specifically, such as Figure 3 As shown: After obtaining the standardized operational data sequence, the internal thermal state distribution relationship from the core center to the outer shell is established by combining the core structure parameters, radial thermal resistance distribution and shell boundary conditions of the target monitoring unit. The internal thermal state distribution relationship is used to characterize the heat accumulation characteristics, radial transfer characteristics and shell boundary constraint characteristics of different radial positions of the core, thereby providing structural parameters and boundary parameters for subsequent continuous updates of the internal thermal state.

[0079] The core structure parameters include at least the core inner diameter, core outer diameter, core bundle height, electrode thickness, diaphragm thickness, effective number of winding layers, shell thickness, and the contact interface state between the core bundle and the shell. For cylindrical supercapacitors, the core inner diameter can be selected from 4mm to 15mm, the core outer diameter from 15mm to 40mm, and the core bundle height from 30mm to 120mm. These ranges are determined based on the geometric dimensions of common small and medium-sized cylindrical supercapacitors and can cover the core structure range of commonly used single-unit products. For square wound supercapacitors, the shortest heat conduction path from the core center to the inner wall of the shell is converted into an equivalent radial length, which can be selected from 5mm to 20mm. These ranges are determined based on the square core bundle thickness, the shell cavity dimensions, and the core center position distribution. Electrode thickness... The thickness can be selected from 20μm to 150μm, and the membrane thickness can be selected from 10μm to 40μm. The above range is determined based on the structural thickness of the current collector, electrode active material coating, and membrane material commonly used in double-layer supercapacitors. The effective number of winding layers is determined based on the capacity level and core size of the target device, and is usually selected from 100 to 500 layers. The above range is determined based on the actual layer distribution of common high-capacity wound supercapacitors. The core structure parameters are preferentially based on design drawings, factory specifications, or sample measurement results. When complete structural data is lacking, the corresponding capacity level and size level parameter combination can be selected from the preset structural parameter table for calculation based on the rated capacity, external dimensions, and weight. Among them, the core outer diameter and core package height are determined first based on the external dimensions, and then the number of layers and thickness level are determined based on the rated capacity and weight.

[0080] To enable the internal thermal distribution to be directly used in subsequent processing, the heat conduction path from the core center to the outer shell is discretized into multiple radial levels. The number of radial levels can be selected from 3 to 8, for example, 4 or 5. This range is determined based on the equivalent radial length, the complexity of the core structure, and online processing resources. When the equivalent radial length is less than 8 mm, the radial levels can be selected from 3 to 4; when the equivalent radial length is 8 mm to 15 mm, the radial levels can be selected from 4 to 6; when the equivalent radial length is greater than 15 mm, the radial levels can be selected from 5 to... Eight layers; fewer than three layers are insufficient to distinguish the central heat accumulation zone and the boundary heat exchange zone; more than eight layers will significantly increase the online processing burden while having limited benefits for engineering applications; when four layers are used, the radial direction can be divided into a central layer, an inner transition layer, an outer transition layer, and a boundary layer; when five layers are used, an intermediate buffer layer can be added between the central layer and the boundary layer; the central layer corresponds to the deep region where heat is most easily accumulated, the inner and outer transition layers correspond to the transition regions where heat is transferred from the center of the core to the outside, and the boundary layer corresponds to the region where heat is thermally coupled with the inner surface of the shell;

[0081] After completing the hierarchical division, node position parameters, node thickness parameters, node thermal capacity parameters, inter-node thermal resistance parameters, and boundary coupling parameters are established for each radial level. The node position parameters are used to represent the relative position of each layer in the direction from the core center to the inner wall of the shell. The node thickness parameters are determined according to the total radial length and the number of levels. The node thermal capacity parameters are obtained by first determining the total equivalent thermal capacity of the target monitoring unit, and then distributing it to each layer according to the material mass ratio or volume ratio of each layer. The total equivalent thermal capacity of the target monitoring unit can be selected from 120J / K to 380J / K. The above range is determined based on the mass and specific heat capacity conversion results of the metal foil, porous active material, electrolyte, diaphragm, and the shell part involved in heat exchange in the target monitoring unit. When the capacity of the target monitoring unit is 500F to 3000F and the unit mass is 0.3kg to 1.5kg, the total equivalent thermal capacity is usually within the above range. The node thermal capacity of each layer is obtained by distributing the total equivalent thermal capacity according to the material composition ratio of the corresponding layer. Therefore, the node thermal capacity values ​​of different layers are different.

[0082] The inter-node thermal resistance parameters are obtained by first determining the total equivalent radial thermal resistance from the core center to the outer shell, and then distributing it to adjacent layers according to the thickness of each layer, the thermal conductivity of the material, and the interface contact state. The total equivalent radial thermal resistance can be selected from 3.5K / W to 7.5K / W. The above range is determined comprehensively based on the equivalent radial length of the core, the effective number of winding layers, the thermal conductivity of the diaphragm, the degree of electrolyte wetting, and the contact state between the core and the shell. When the equivalent radial length is 5mm to 20mm and the effective number of winding layers is 100 to 500, the total equivalent radial thermal resistance is usually within the above range. When the thermal resistance between adjacent layers is distributed according to the radial position, it can be set to a distribution pattern that gradually changes from the center to the outside according to the tightness of the structure and the state of immersion, so as to reflect the difference in thermal conductivity of different areas of the core. For structures with high core tightness and sufficient immersion, the thermal resistance near the outer layers can be appropriately reduced. For structures with low core tightness or insufficient local immersion, the thermal resistance near the center layers can be appropriately increased.

[0083] The enclosure boundary conditions are jointly determined by the enclosure temperature data and environmental conditions in the standardized operating data sequence. The enclosure boundary conditions include at least the following parameters: enclosure temperature, ambient temperature, heat dissipation medium state, cooling level, heat transfer boundary level, and installation posture correction state. If the operating environment is natural air cooling, the natural convection boundary between the enclosure surface and the air is used. If the operating environment is forced air cooling or liquid cooling, the corresponding heat dissipation medium state and cooling level are written according to the operating status of the heat dissipation device. The heat transfer boundary level is used to characterize the enclosure's ability to dissipate heat outward and can be divided into normal level, confined level, and enhanced level. The installation posture correction state is used to characterize the influence of the device installation direction on natural convection heat transfer and local contact state.

[0084] The establishment of the internal thermal state distribution relationship includes basic structure mapping and boundary state correction. Basic structure mapping is determined based on the core geometry, layer division results, heat capacity of each layer node, and inter-node thermal resistance, used to describe the heat conduction path from the core center to the outer shell under standard boundary conditions. Boundary state correction adjusts the boundary heat dissipation capacity based on the outer shell temperature, ambient temperature, cooling intervention status, and installation posture. When the outer shell temperature is continuously 20°C higher than the ambient temperature for 5 to 30 seconds, the heat transfer boundary level can be switched from the conventional level to the restricted level. This temperature difference threshold and duration range are determined based on the common steady-state temperature difference range between the shell surface and the environment under natural or forced air cooling conditions. When the outer shell temperature is continuously 20°C higher than the ambient temperature within the above time window, it usually indicates a decrease in boundary heat dissipation capacity or increased internal heat accumulation. When the outer shell temperature drops by 2°C to 5°C within a 1 to 5-second time window after external cooling has been activated, the heat transfer boundary level can be switched to the enhanced level. This time window and temperature drop range are determined based on the shell thermal inertia, the response speed of common air-cooling or liquid-cooling systems, and surface temperature drop characteristics.

[0085] To improve parameter reproducibility, the total equivalent heat capacity is primarily calculated based on the bill of materials, mass distribution, and specific heat capacity. When complete material information is lacking, it can be calculated using preset heat capacity parameter tables for devices of the same capacity and mass level. The total equivalent radial thermal resistance is primarily determined based on the thermal response calibration data of the target device under typical operating conditions. When calibration data is unavailable, it can be selected from the preset thermal resistance parameter set of devices with similar size and capacity levels, and corrected based on the actual thermal response during subsequent operation. Boundary condition parameters primarily use real-time shell temperature, ambient temperature, and heat dissipation status. When boundary condition data is missing, a conservative boundary level is automatically adopted to avoid underestimating the risk of internal heat accumulation.

[0086] After completion, an internal thermal state distribution dataset is generated. This dataset includes at least the following fields: layer node number, node location, node thickness, node thermal capacity, inter-node thermal resistance, heat transfer boundary level, boundary correction status, and current update time. This dataset can be directly used as the input basis for subsequent continuous updates of the internal thermal state. To meet online control requirements, the processing time from receiving the standardized operating data sequence to generating a complete internal thermal state distribution relationship can be selected to be no more than 500 μs. This time range is determined based on the high-frequency sampling period, the controller's processing capacity, and the reserved time required for subsequent status updates. If the device model is fixed and the core structure parameters remain unchanged, the basic structure mapping can be generated once during the power-on initialization phase. During operation, only the boundary state correction part is refreshed according to the shell temperature sampling period or the thermal state update period to reduce the online processing burden. If structural parameters are missing, the layer division cannot be completed, or the boundary conditions are invalid, a preset internal thermal state distribution relationship of no less than 3 layers is selected according to the capacity level and external dimensions, and a conservative operation flag is written to ensure that subsequent processing can continue.

[0087] Specifically, such as Figure 4 As shown: After obtaining the standardized operating data sequence and the internal thermal state distribution dataset, the internal thermal state of different radial positions of the core is continuously refreshed according to a fixed update time. The internal thermal state includes at least the current temperature value, temperature change rate, heat accumulation level, and outward heat transfer status level of each radial level. The current temperature value is used to characterize the temperature estimate of the corresponding level within the current update time, in degrees Celsius. The temperature change rate is used to characterize the temperature change trend between adjacent update times, in °C / s. The heat accumulation level is used to characterize the degree of heat accumulation of the corresponding level, which can be divided into stable level, rising level, warning level, and high-risk level. The outward heat transfer status level is used to characterize the smoothness of heat transfer to the outer layer, which can be divided into normal, restricted, and severely restricted. The discrete level form is adopted to facilitate subsequent pulse timing adjustment and heat dissipation intervention control based on different internal thermal states.

[0088] The internal thermal status is updated according to a fixed update time; the update time can be selected from 1ms to 5ms, preferably 2ms; this time range is determined comprehensively based on the high-frequency sampling cycle of terminal voltage and bidirectional current, the accumulation rate of heat in the deep core in the early stage, and the online processing capability of the controller; when the update time is greater than 5ms, it is easy to delay the response to high-frequency pulse heat input; when the update time is less than 1ms, it will significantly increase the online processing burden, while the improvement of status discrimination accuracy is limited; when the high-frequency sampling frequency is 10kHz to 12kHz, the update time can correspond to 10 to 24 sampling cycles, thus balancing data coverage and real-time processing; within each update time, the most recent data window before the current update time is read, and combined with the current shell temperature, heat transfer boundary level, and internal thermal status distribution relationship, the heat input intensity, outward heat transfer status, and internal thermal status level of each radial layer are updated; the data read includes at least the fields of terminal voltage, bidirectional current, shell temperature, heat transfer boundary level, and internal thermal status of each layer at the previous update time;

[0089] To ensure the reproducibility of heat input discrimination, a heat load discrimination window is set. The heat load discrimination window can be selected from 20ms to 100ms, preferably 50ms. This time range is determined comprehensively based on the common duration of high-frequency pulse grouping, the accumulation process of heat deep in the core in a short time, and the feedforward required for online judgment. When the window is less than 20ms, it is easily affected by the occasional fluctuation of a single pulse. When the window is greater than 100ms, it will weaken the sensitivity to rapid changes in heat input. Within the heat load discrimination window, the effective time percentage of the absolute value of bidirectional current exceeding the rated continuous current multiple threshold is statistically analyzed. The rated continuous current multiple threshold can be selected from 3 to 5 times, preferably 4 times. This threshold is set comprehensively based on the temperature rise curve of the target device under rated continuous current and the allowable pulse range under high-rate transient conditions. When the pulse current exceeds 4 times the rated continuous current, the accumulation of Joule heat and polarization heat is usually significantly enhanced, which can be used as a boundary to distinguish between conventional pulses and high heat input pulses.

[0090] If, within the heat load discrimination window, the effective time percentage for which the absolute value of the bidirectional current exceeds four times the rated continuous current is higher than 40%, the current window is determined to have entered a medium-continuous heat input state, and the heat accumulation level of the central layer is increased by at least one level. This 40% time percentage is used to distinguish between short-term, intermittent large currents and continuous pulse heat inputs. If this percentage is higher than 70%, and the terminal voltage fluctuation amplitude reaches 5% to 10% of the rated voltage, the current window is determined to have entered a strong-continuous heat input state, and the heat accumulation levels of the central layer and the inner transition layer are increased simultaneously. This 70% time percentage is used to identify high-density pulse continuous injection states, and the fluctuation amplitude of the rated voltage of 5% to 10% is used to identify high-pulse operating intervals accompanied by significant energy exchange. The range of this terminal voltage fluctuation amplitude is determined based on the common range of transient deviations of the terminal voltage from the rated voltage level during high-rate pulse charging and discharging. By combining the rate condition, time percentage condition, and voltage fluctuation condition, the misjudgment of the overall heat input intensity based solely on a single current peak value can be avoided.

[0091] After completing the heat load determination, the heat input weight and heat transfer weight of different radial levels are adjusted according to the current direction, pulse continuity, and pulse gap characteristics in the current data window. If a large-amplitude unidirectional charging pulse is detected in multiple consecutive high-frequency time slices, or a pulse group with repeated high-frequency polarity switching is detected, the heat input weight of the central layer and the inner transition layer is increased. This adjustment is determined based on the characteristic that the deep layers of the core are more prone to local heat accumulation during high-density charging or rapid polarity alternation. If continuous discharge is detected and the pulse gap is obvious, the heat transfer weight from the central layer to the outer transition layer is increased. This adjustment is determined based on the characteristic that deep heat is more easily released radially outward during continuous discharge. By dynamically adjusting the heat input weight and heat transfer weight, the current temperature value and outward heat transfer state level of different radial levels are updated, thereby forming a dynamic internal thermal state corresponding to the operating data.

[0092] To ensure the reproducibility of temperature rise trend discrimination, a temperature change rate threshold is set. A temperature change rate exceeding 2℃ / s in the central layer is considered a rapid rise, and exceeding 5℃ / s is considered a sharp rise. The outer transition layer and boundary layer, affected by the shell's heat capacity and boundary heat transfer, experience relatively slower temperature rises; their rapid and sharp rise thresholds can be set to 1℃ / s and 3℃ / s, respectively. The inner transition layer can be discriminated using a threshold range between the central and outer transition layers. These temperature change rate thresholds are preset and calibrated based on radial layer position, layer thickness, update time, and differences in thermal inertia between layers. The central layer threshold is higher than the outer layer threshold to reflect the difference in thermal response between deeper layers and outer layers. For devices with larger equivalent radial lengths or layer thicknesses, fine-tuning can be made in the same direction based on the above thresholds to maintain consistency in discrimination between different devices.

[0093] The internal thermal status update adopts a status inheritance mechanism from the previous update time. When each layer refreshes its status at the current update time, in addition to reading the terminal voltage, bidirectional current, and shell temperature within the current window, it also simultaneously reads the current temperature value, heat accumulation level, and outward heat transfer status level corresponding to the previous update time. If the central layer was already at the alert level at the previous update time, and the heat input within the current window is still higher than the boundary heat dissipation capacity, the central layer will not directly drop back to the stable level, but will at least maintain the alert level or rise to the high-risk level. If the boundary layer was at a high temperature at the previous update time and the central layer has shown a significant drop, it indicates that the deep heat was being transferred outward in the early stage. At this time, the state will not be misjudged as a continuous deterioration of the deep layer, but the outward heat transfer status level of the corresponding layer will be marked as normal transfer. Through this inheritance mechanism, the impact of instantaneous fluctuations within a single update time on the status determination can be reduced, and the resulting control false triggering can be suppressed.

[0094] To improve feasibility under abnormal conditions, abnormal and boundary handling rules are set. When the proportion of invalid frames in the terminal voltage or bidirectional current exceeds 30% within an update time, the internal thermal state update is not stopped. Instead, the average heat input trend of the most recent consecutive valid windows is called to put the internal thermal state into a conservative update mode. The 30% threshold is determined based on the sampling validity requirements and the tolerance range of abnormal data. It is used to distinguish between occasional sampling loss and obvious data distortion, maintaining update continuity while avoiding complete reliance on abnormal data. When the shell temperature probe fails, the boundary layer is no longer updated according to the measured temperature. Instead, it is conservatively estimated based on the most recent valid shell temperature value, the current heat transfer boundary level, and the heat dissipation intervention status. If the shell temperature is higher than 88°C, or the central layer is in a high-risk level for 3 to 10 consecutive update times and the temperature change rate is still positive, a thermal runaway precursor flag is written. The 88°C temperature threshold is determined based on the high-temperature abnormal warning temperature zone of the organic electrolyte supercapacitor and the characteristic that the shell temperature measurement has a hysteresis margin relative to the deep core.

[0095] The number of consecutive cycles is determined based on the update time. When the update time is 1ms, 8 to 10 cycles can be selected; when the update time is 2ms, 5 cycles are preferred; and when the update time is 5ms, 3 to 5 cycles can be selected to distinguish between short-term fluctuations and continuous deterioration. If the environment is at extremely low temperatures, such as when the shell temperature is below -35°C, the thermal accumulation response of the central layer will be amplified due to the increased viscosity and internal resistance of the electrolyte. In this case, the sensitivity of the increase in the thermal accumulation level of the central layer is increased. The -35°C threshold is determined based on the low-temperature operating environment level of the power equipment and the low-temperature flow characteristics of the organic electrolyte to avoid underestimating the hidden heating under low-temperature high-current conditions.

[0096] After the above refresh is completed, a dynamic internal thermal state sequence is formed. For each update time, the output results include at least the current timestamp, the current temperature value of each radial level, the temperature change rate of each layer, the heat accumulation level of each layer, the outward heat transfer state level of each layer, the risk indicator of the central layer, and the global thermal state level. The global thermal state level is determined by the central layer and the inner transition layer and can be divided into normal state, early warning state, high-risk state, and protection state. This output result can be directly used as the basis for subsequent determination of the charging and discharging pulse conduction period, interval period, polarity alternation sequence, and external heat dissipation intervention sequence.

[0097] To meet online control requirements, the processing time for a single internal thermal state update can be selected to be no more than 1ms, preferably no more than 800μs. This time range is determined based on the update time, the controller's processing capacity, and the timing reservation for subsequent pulse regulation. If the processing time for a single update exceeds the above range, the thermal load discrimination window length can be actively reduced or the boundary state correction frequency can be simplified, but the thermal load discrimination window should not be less than 20ms, and the central layer state update should not be canceled, thereby maintaining the priority of deep hot spot identification.

[0098] Specifically, such as Figure 5 As shown: After obtaining the dynamic internal thermal state sequence, a pulse control sequence is generated based on the current temperature value of the central layer, the temperature change rate of the central layer, the thermal accumulation level of the inner transition layer, the temperature value of the boundary layer, the temperature change rate of the boundary layer, the global thermal state level, the risk indicator of the central layer, and the precursor indicator of thermal runaway. The pulse control sequence includes at least the following fields: target conduction period, target interval period, target polarity arrangement indicator, external heat dissipation intervention level, current operating state, and backoff lockout indicator. The current operating state can be divided into normal operating state, pre-control state, enhanced heat dissipation state, and protection switching preparation state. A state machine is used because different internal thermal states correspond to different pulse organization methods and heat dissipation intervention methods. This form is beneficial for maintaining the consistency of entry conditions, backoff conditions, and protection preconditions.

[0099] When the global thermal state level is normal and the temperature change rate of the central layer is less than 2℃ / s, the original target pulse structure is maintained, and the current conduction period and interval period are recorded as the baseline control parameters. The temperature change rate of the central layer is taken as 2℃ / s as the boundary between the normal operation state and the pre-control state. This value is consistent with the threshold for judging rapid temperature rise of the central layer in the previous internal thermal state update. When the temperature change rate of the central layer reaches or exceeds this threshold, it indicates that the deep heat accumulation of the core has entered the range that requires pre-intervention. If the global thermal state level rises to the warning state, or the temperature change rate of the central layer reaches 2℃ / s or above, the system switches to the pre-control state. The state switching adopts a judgment method that enters when any condition is met, so as to intervene in the control in advance before the local rapid temperature rise of the central layer is fully conducted to the global thermal state level.

[0100] After entering the pre-regulation state, without changing the target peak value of a single pulse, priority is given to adjusting the conduction period and interval period. The conduction period is defined as the time interval during which a charging or discharging pulse can be continuously output, in milliseconds. The interval period is defined as the gap time between adjacent high-heat input pulses, in milliseconds. In the pre-regulation state, the conduction period for a single high-heat input can be selected from 5ms to 30ms, preferably from 10ms to 20ms. This range is determined based on the shortest pulse duration required for effective power support and the longest continuous conduction time that does not cause continuous deterioration of heat accumulation in the central layer under the pre-regulation state. When the conduction period is less than 5ms, the single pulse is effective. Insufficient energy delivery can easily reduce the effectiveness of power support; when the conduction period is longer than 30ms, continuous heat input will significantly increase deep heat accumulation; the interval period in the pre-controlled state can be selected from 15ms to 45ms, preferably 20ms to 30ms; this range is determined based on the ion concentration recovery time of the porous electrode interface and the minimum interval required for heat release; when the interval period is less than 10ms, the effect on ion concentration recovery and heat release is limited; when the interval period is longer than 50ms, the ability to support high power continuously is easily weakened; by shortening the duration of continuous high heat input and inserting a recovery interval, a heat release interval can be formed between adjacent high heat input pulses;

[0101] When there is room for bidirectional power scheduling, the polarity alternation sequence is optimized; the polarity arrangement flag is used to characterize the charging and discharging sequence of adjacent pulses; this scheme does not adopt a random switching method, but determines the priority order based on the temperature change trend of the central layer and the heat accumulation level of the inner transition layer; when the central layer continues to heat up and the heat accumulation level of the inner transition layer reaches the warning level or above, if the controlled object allows the charging and discharging pulses to alternate adjacently, the discharging pulse is preferentially arranged after the high heat input charging pulse; this arrangement is determined based on the state characteristics of the high-rate discharge stage which are conducive to reducing the subsequent heat accumulation in the deep layers; if the polarity alternation conditions are not available, the original polarity direction is kept unchanged, and heat release is preferentially achieved by extending the interval period;

[0102] If the global thermal status level rises to a high-risk state, or the central layer risk flag is activated, the system switches to enhanced heat dissipation mode. In enhanced heat dissipation mode, while maintaining the target load requirements, the conduction period is further narrowed and the interval period is expanded. The conduction period can be selected from 5ms to 20ms, determined based on the allowed continuous heat input duration within the high-risk range. When the conduction period exceeds 20ms, the accumulation rate of heat in the central layer within a single pulse is likely to exceed the outward heat transfer rate. The interval period can be selected from 50ms to 200ms, with the lower limit of 50ms determined based on the effective release time required for heat transfer from the central layer to the outer layer, and the upper limit of 200ms determined based on the target load's tolerance time to transient power interruptions. When the interval period exceeds 200ms, the continuous power support capability will significantly decrease; therefore, 200ms is set as the upper limit for the interval in enhanced heat dissipation mode. Through these adjustments, more time can be allocated for heat transfer from the central layer to the outer transition layer and boundary layer.

[0103] The external heat dissipation intervention sequence is based on changes in the inner layer's thermal state, not solely on exceeding the outer shell temperature limit. External heat dissipation intervention levels are categorized into standby, pre-start, and execution levels. The standby level indicates maintaining the current heat dissipation readiness state, the pre-start level indicates entering a rapid start-up state in advance, and the execution level indicates issuing an external heat dissipation activation request. When the central layer temperature change rate exceeds 2℃ / s for three consecutive update intervals and the boundary layer temperature change rate is below 0.5℃ / s, the external heat dissipation intervention level can be switched from standby to pre-start. The condition of three consecutive update intervals is determined based on the short-term sustainability requirement of rapid central layer temperature rise. The boundary layer temperature change rate is set at 0.5℃ / s as an auxiliary threshold because this value is below the boundary layer rapid temperature rise detection threshold. This can be used to characterize the early thermal mismatch range where the deep layer has rapidly heated up but the boundary layer has not yet entered a significant heating state. If the current temperature of the central layer reaches the preset warning temperature, the external heat dissipation intervention level will be switched from the pre-start level to the execution level. The preset warning temperature can be selected from 55℃ to 60℃. This range is determined based on the warning temperature zone during the continuous heating phase of the central layer and the temperature range that can still play a pre-suppression role after the external heat dissipation system is started. When the heat dissipation system starts up slowly, the lower limit of this range can be used first. When the heat dissipation system responds quickly, the upper limit of this range can be used first. If the risk flag of the central layer has been activated, the system can be directly switched to the execution level regardless of whether the current temperature of the central layer has reached the above range. After entering the execution level, an external heat dissipation start request will be issued simultaneously while the subsequent pulse timing adjustment is implemented.

[0104] If the current temperature of the central layer reaches 90% to 98% of the preset limit temperature, or reaches the range of 65°C to 75°C, or the thermal runaway precursor flag is activated, the system enters the protection switching preparation state. The preset limit temperature is determined based on the target device's allowable safe temperature of the central layer and the housing temperature measurement hysteresis margin. When the upper limit of the target device's safe temperature of the central layer is known, a relative threshold of 90% to 98% is preferentially used as the pre-switching condition. When the upper limit of the safe temperature of the central layer cannot be directly given, an absolute pre-protection temperature range of 65°C to 75°C can be used. The above relative and absolute thresholds are set to reserve execution time for subsequent graded power limiting control and disconnection protection before the actual protection action threshold is reached. After entering the protection switching preparation state, the state machine maintains the contraction of the conduction period, the expansion of the interval period, and the maintenance of the external heat dissipation execution stage.

[0105] To avoid control oscillations, a state machine rollback rule is set. The pre-regulation state can only roll back to the normal operating state when the temperature change rate of the central layer is below 0.5℃ / s for 3 to 10 consecutive update times, and the heat accumulation level upgrades or the warning level decreases. The 0.5℃ / s threshold is determined based on the requirement for the central layer to transition from a rapid heating phase to a stable and moderate phase. The number of consecutive cycles is determined by the update time length: 8 to 10 cycles are selectable when the update time is 1ms, 5 cycles are preferred when the update time is 2ms, and 3 to 5 cycles are selectable when the update time is 5ms. Enhanced heat dissipation state. The system can only revert to the pre-regulation state when the current temperature of the central layer drops below the preset warning temperature and remains below it for at least 5 seconds. This 5-second holding time is determined based on the thermal inertia of the air-cooled or liquid-cooled system and the heat dissipation stabilization time. Before actual protection action is initiated, the protection switching preparation state can only exit when the continuous drop condition is met. The continuous drop condition can be selected as the central layer temperature decreasing for 3 to 10 consecutive update times, with the temperature change rate turning negative, and the heat accumulation level decreasing by at least one level. This condition is used to distinguish between short-term fluctuations and the actual drop process, thereby avoiding frequent entry and exit from the protection preparation state.

[0106] When the controller's computing power is limited, priority is given to ensuring the state machine branch judgment, the output of the target conduction period, and the output of the target interval period, and then the external heat dissipation intervention level refinement and polarity arrangement optimization are performed; if there are missing fields in the dynamic internal thermal state sequence, the state machine directly enters a conservative branch of no less than the pre-controlled state and does not continue to maintain the normal operating state, so as to avoid underestimating the thermal risk when the input state is incomplete.

[0107] After completing the above determination, a pulse control sequence table is formed. This sequence table is based on control cycles and includes at least the following fields: target conduction period, target interval period, target polarity arrangement flag, external heat dissipation intervention level, current operating status, and backoff lock flag. This sequence table is used for subsequent timing adjustment of charging and discharging pulses and provides prerequisites for graded power limiting control and disconnection protection. To meet online control requirements, the single processing time from receiving the dynamic internal thermal state sequence to generating the pulse control sequence table can be selected to be no more than 500μs, preferably no more than 300μs. This time range is determined jointly based on the update time, controller processing capacity, and timing reservation of subsequent execution links. When the processing time exceeds the above range, priority is given to retaining the state switching judgment and basic conduction interval reorganization, without canceling critical safety branches.

[0108] Specifically, such as Figure 6 As shown: After obtaining the pulse control sequence table, the timing of the charge and discharge pulses is adjusted according to the control cycle. Combined with the latest global thermal state level, central layer risk indicator, thermal runaway precursor indicator, and abnormal operating status indicator, graded power limiting control and disconnection protection are implemented. The controlled object is the charge and discharge control link of the supercapacitor. The execution content includes at least pulse conduction release, pulse interval maintenance, external heat dissipation execution indicator output, power limiting level switching, disconnection protection triggering, lockout protection maintenance, and recovery condition verification. The control cycle can be selected from 1ms to 10ms, preferably 2ms to 5ms. This range is determined based on the preceding internal thermal state update time, the execution link output delay, and the load side's requirements for pulse response real-time performance. When it is necessary to follow the internal thermal state cycle by cycle, the control cycle can preferably be 2ms to 5ms. When the execution link uses merged output and the load allows for batch updates in multiple cycles, the control cycle can be relaxed to 10ms. When the control cycle is less than 1ms, it will significantly increase the burden on the execution link, while having limited improvement on control accuracy.

[0109] At the start of each control cycle, the target conduction period, target interval period, target polarity arrangement flag, external heat dissipation intervention level, current operating status, and backoff lock flag fields in the current pulse control sequence table are read and combined with the latest dynamic internal thermal state sequence for execution verification. If the current operating status is normal operating status, the charging pulse or discharging pulse is released according to the original target rhythm. If the current operating status is pre-control status, the interval period is inserted between adjacent high-heat input pulses according to the sequence table, while keeping the single pulse target peak value unchanged. If the current operating status is enhanced heat dissipation status, the external heat dissipation execution flag is output while performing conduction period compression and interval period expansion. If the current operating status is protection switching preparation status, the power limit boundary and disconnection protection preparation conditions are preset while performing conduction period contraction, interval period expansion, and external heat dissipation execution level maintenance.

[0110] To ensure the reproducibility of power limiting control, it is divided into at least three levels. The first power limiting level corresponds to the state where the current temperature of the central layer reaches the warning upper limit but has not reached the protection temperature range, or the state where the global thermal status is in a high-risk state and the thermal runaway precursor flag has not yet been activated. The warning upper limit is preferably the preset warning temperature upper limit determined in the previous step, i.e., the upper limit value in the range of 55°C to 60°C. When the preset warning temperature is 60°C, the central layer temperature range corresponding to the first power limiting level can be 60°C to 65°C. After entering the first power limiting level, the upper limit of the cumulative conduction time allowed is shortened to 60% to 80% of the upper limit of the cumulative conduction time corresponding to the current pulse control sequence table, preferably 70%. This ratio range is determined based on the control objective of reducing continuous heat input by 20% to 40% while ensuring basic power support. Furthermore, the average pulse duty cycle is limited to 20% to 50%, preferably 30% to 40%. This range is determined based on the combination relationship between the conduction period and the interval period in the pre-control state and the minimum pulse density required to maintain continuous operation.

[0111] The second power limit level corresponds to a further increase in the current temperature of the central layer, or a sustained positive rate of temperature change in the central layer and a failure of the boundary layer to effectively follow the temperature increase. A further increase in the current temperature of the central layer can be defined as 5°C to 10°C higher than the upper limit of the first power limit level. For example, when the upper limit of the first power limit level is 65°C, the central layer temperature range corresponding to the second power limit level can be 65°C to 75°C. Failure of the boundary layer to effectively follow the temperature increase can be defined as a sustained positive rate of temperature change in the central layer and a boundary layer temperature change rate lower than 0.5°C / s for 3 to 10 consecutive control cycles, or a boundary layer temperature change rate lower than 0.5°C / s within the same observation time window. The temperature rise is less than 1°C to 2°C; the observation time window can be selected as the total duration corresponding to 3 to 10 consecutive control cycles; the range of 0.5°C / s and 1°C to 2°C is determined based on the boundary layer thermal inertia and the threshold for significant temperature rise in the boundary layer; after entering the second power limit level, in addition to continuing to shorten the cumulative conduction time, the single target peak holding time is also limited to 5ms to 15ms; this range is determined based on the allowable single pulse duration in the high thermal risk zone and the conduction period constraint under the previous enhanced heat dissipation state; and the average duty cycle is further narrowed to 10% to 30%, preferably 15% to 25%, to further reduce the continuous heat input density;

[0112] The third power limit level corresponds to the state where the thermal runaway precursor flag is activated but the immediate disconnection condition has not yet been met; after entering the third power limit level, it switches to a low duty cycle retention mode; the average duty cycle in the low duty cycle retention mode can be selected from 5% to 15%, preferably 10%; this range is determined by the minimum pulse density required to maintain the minimum continuous operation capability and the constraint of avoiding the central layer from continuing to heat up rapidly; at this time, the single conduction period can be compressed to 5ms to 10ms, and the interval period is maintained near the upper limit of the enhanced heat dissipation state or further extended, but does not exceed the upper limit of the power interruption allowed by the target load;

[0113] The disconnection protection conditions include at least temperature over-limit conditions, voltage abnormal conditions, and execution abnormal conditions. If the current temperature value of the center layer reaches the disconnection threshold, and the temperature change rate of the center layer remains greater than 0 for 3 to 10 consecutive update times, then disconnection protection is triggered. The disconnection threshold is primarily determined based on the upper limit of the safe center layer temperature allowed by the target device. When the upper limit of the safe center layer temperature is known, 95% to 100% of this upper limit can be used as the disconnection threshold. This percentage range is used to cover the execution delay of the protection action and the hysteresis margin of the housing temperature measurement relative to the depth of the core. When the upper limit of the safe center layer temperature cannot be directly obtained, a preset value can be selected within the range of 75°C to 85°C. The 75°C to 85°C range is determined based on the organic electrolyte. The requirements for determining the deep high-risk temperature zone of the supercapacitor, the temperature measurement hysteresis margin of the shell, and the pre-action requirements for protection actions are as follows: If an abnormal voltage drop occurs at the terminal and is accompanied by the activation of the risk indicator in the central layer, it is determined that there may be an internal short circuit, insulation failure, or severe energy release abnormality, and the disconnection protection is triggered; An abnormal voltage drop at the terminal can be defined as a drop in terminal voltage exceeding 20% ​​to 40% of the rated voltage within one to three consecutive control cycles; This range is determined based on the voltage fluctuation amplitude under normal pulse conditions and the rapid drop characteristics under abnormal fault conditions; If the abnormal operation status indicator shows severe sampling distortion, control command mismatch, or external heat dissipation failure, and the global thermal status level is already in the protection state, the disconnection protection is directly initiated;

[0114] After the protection is disconnected, subsequent charging and discharging pulses are prohibited from being released, and the control status is locked to the protection status. If an abnormal terminal voltage or a continued rise in the central layer temperature is detected after the protection is disconnected, the protection status is upgraded to the locked protection status. In the locked protection status, automatic recovery is prohibited, and manual reset is only allowed after the fault is investigated. Before manual reset, at least the following conditions must be checked: the abnormal terminal voltage flag is cleared, the central layer temperature is below the upper limit of the recovery temperature, the central layer temperature change rate has fallen back to the stable range, the abnormal operating status flag is cleared, and the external heat dissipation is in normal operation. After manual reset, the system does not directly enter the normal operating state, but first enters the first power limit level, and then gradually retreats according to subsequent thermal state changes.

[0115] The recovery conditions are used to define the boundary for retreating from the protection state or power-limited state to a lower-risk state. The automatic recovery conditions can be set as follows: the current temperature of the center layer is continuously maintained below 40℃ for no less than 10 minutes, and the absolute value of the temperature change rate of the center layer does not exceed 0.2℃ / s to 0.5℃ / s, preferably 0.3℃ / s, while the abnormal operation status flag is released. The 40℃ threshold is determined based on the requirement that the temperature is below the warning temperature zone and close to the normal temperature safe recovery zone. The 10-minute duration is determined based on the time required for the core, shell, and boundary heat dissipation system to reach a stable temperature drop. The absolute value of the temperature change rate not exceeding 0.2℃ / s to 0.5℃ / s is used to characterize that the temperature has approached stability and is no longer in a continuous heating or rapid cooling process. For liquid cooling systems, the lower limit of the range can be used first, and for air cooling systems, the upper limit of the range can be used first. After the automatic recovery conditions are met, the system does not directly return to the normal operation state, but first enters the first power-limited level, and then gradually recovers to the pre-controlled state or normal operation state according to the thermal state change.

[0116] To maintain the closed-loop nature of the execution link, feedback adjustment rules are set. If, after the execution timing adjustment, the subsequent internal thermal status update results show that the central layer temperature change rate does not decrease within 2 to 5 consecutive control cycles, or the central layer heat accumulation level does not decrease within 2 to 5 consecutive control cycles, it is determined that the current combination of the conduction period and the interval period is insufficient to suppress heat accumulation, and an automatic request to extend the interval period or increase the external heat dissipation intervention level is made in the next control cycle. If, after the execution of graded power limiting control, the boundary layer temperature change rate increases while the central layer temperature decreases, it is determined that deep heat is being released outward, and the current power limiting level is maintained until the release process is completed. If the voltage at the back end of the protection execution is still abnormal or the temperature rise continues to rise, the locked protection state is maintained, and the automatic recovery process is not entered.

[0117] To meet resource constraints, the maximum processing delay for the execution and protection links is specified. The time from reading the pulse control sequence table to completing the output of the execution instruction for this cycle can be set to no more than 300μs. This time range is determined based on the control cycle, execution interface delay, and the reserved time for subsequent protection links. The time from detecting the establishment of the disconnection protection condition to outputting the protection instruction can be set to no more than 2ms, preferably no more than 1ms. This time range is determined based on the protection action pre-requisites and the controller's worst-case preemptive response capability. If the controller's computing power is limited, priority is given to ensuring the execution of state switching judgment, target conduction period output, target interval period output, and disconnection protection path execution, followed by external heat dissipation intervention level refinement and polarity arrangement optimization, without canceling critical safety branches.

[0118] After the above execution is completed, an execution status record and operational status feedback information are generated. The execution status record includes at least the following fields: whether the timing adjustment of the current control cycle has been completed, the current power limit level, whether the disconnection protection has been triggered, and whether the locked protection state has been entered. The operational status feedback information includes at least the following fields: timestamp, reason for state switching, central layer temperature, boundary layer temperature, current power limit level, external heat dissipation level, and recovery status. The above information is used for subsequent verification, parameter correction, and the next round of internal thermal status update and pulse control sequence recalculation.

[0119] Example 2: Based on Example 1, the specific application process of a supercapacitor temperature rise suppression optimization method is further explained:

[0120] Taking a supercapacitor module used in wind power transient power support scenarios as an example, the target monitoring object is the target monitoring unit in the module, which is composed of multiple supercapacitor cells with a rated voltage of 2.7V. The target monitoring unit is equipped with a terminal voltage sampling channel, a bidirectional current sampling channel, and a shell temperature sampling channel. The controller continuously monitors the target monitoring unit according to a unified time reference and adjusts the charging and discharging pulse timing and external heat dissipation intervention in conjunction with the changes in the deep thermal state of the core, thereby maintaining the necessary transient power output capability while suppressing internal temperature rise.

[0121] During operation, the first step is to collect and process operational data. The controller selects the positive and negative terminals of the target monitoring unit, as well as the temperature sampling location on the casing surface, as data acquisition points. Specifically, the terminal voltage sampling circuit has a range of 0V to 3.2V, with an effective operational identification upper limit of 2.9V; the bidirectional current sampling range is -1500A to +1500A; and the casing temperature sampling range is -40℃ to 110℃. The sampling frequency for terminal voltage and bidirectional current is 10kHz, and the sampling frequency for casing temperature is 10Hz. The controller uses the high-frequency sampling time slices of terminal voltage and bidirectional current as a unified time slice and performs time alignment for casing temperature by holding the most recent valid value, thus forming standardized operation. Data frames; each standardized running data frame includes at least the following fields: timestamp, terminal voltage, bidirectional current, casing temperature, data validity flag, and sampling status. If the terminal voltage exceeds the sampling range, a resampling is triggered. If three consecutive frames are still abnormal, a conservative processing flag is written. If the bidirectional current change rate exceeds 60A / μs, it can be identified as spike interference, and adjacent historical valid data is called for slope smoothing correction. If the casing temperature change rate exceeds 15℃ / s, it is identified as a temperature sampling abnormality, and the previous valid temperature value is maintained. After the above processing, the standardized running data frame is written to a circular buffer. The length of the circular buffer can be 8192 frames to support subsequent short-term historical window calls.

[0122] After processing the operational data, the controller establishes the internal thermal state distribution relationship from the core center to the outer shell based on the structural parameters of the target monitoring unit. Taking a cylindrical wound supercapacitor as an example, the core inner diameter can be 8mm, the core outer diameter can be 28mm, the core pack height can be 60mm, the electrode thickness can be 80μm, the separator thickness can be 25μm, and the effective number of winding layers can be 260. The controller discretizes the heat conduction path from the core center to the outer shell into four radial levels: the center layer, the inner transition layer, the outer transition layer, and the boundary layer. Then, the total equivalent heat capacity of the target monitoring unit is determined, for example, it can be 220J / K, and then distributed to each layer according to the mass ratio of each layer's material. The total equivalent radial thermal resistance is determined, for example, 5.2 K / W, and then distributed between adjacent layers according to layer thickness, material thermal conductivity, and interface contact state. The shell boundary conditions are composed of shell temperature, ambient temperature, heat dissipation medium state, and heat transfer boundary level. If the current state is natural air cooling, the heat transfer boundary level is initially set to the conventional level. If the shell temperature is more than 20°C higher than the ambient temperature for 10 seconds, the heat transfer boundary level is switched to the restricted level. If external cooling is activated and the shell temperature drops by 3°C within 3 seconds, the heat transfer boundary level is switched to the enhanced level. Based on the above structural parameters and boundary parameters, an internal thermal state distribution relationship dataset is formed, which serves as the basis for subsequent internal thermal state updates.

[0123] During the internal thermal state update process, the controller continuously refreshes the state of each layer according to a fixed update time; the update time can be 2ms, and the thermal load discrimination window can be 50ms. Within each update time, the controller retrieves the terminal voltage, bidirectional current, and shell temperature data from the most recent 50ms window, and calculates the current temperature value, temperature change rate, heat accumulation level, and outward heat transfer state level of each layer based on the internal thermal state distribution relationship. Assuming that during a certain operating phase, the target monitoring unit, in response to a transient power support request from the wind power generation side, continuously experiences high-rate charging pulses, and within the 50ms window, the effective time percentage where the absolute value of the bidirectional current exceeds four times the rated continuous current reaches 72%, while the terminal voltage fluctuation amplitude reaches 8% of the rated voltage; under these circumstances, the controller determines the current... The window is under a strong and continuous heat input state, and the heat accumulation level of the central layer and the inner transition layer is raised simultaneously. If the temperature change rate of the central layer reaches 3.1℃ / s at this time, the heat accumulation level of the inner transition layer rises to the warning level, while the temperature change rate of the boundary layer is only 0.3℃ / s, it indicates that there has been significant heat accumulation in the deep layers of the core, while the boundary layer has not yet heated up synchronously, and the internal temperature rise shows the characteristics of the deep layers leading the way. If the central layer was already at the warning level in the previous update time, and the heat input in the current window is still higher than the heat dissipation capacity of the boundary layer, the central layer will not fall back to the stable level, but will continue to maintain the warning level or rise to the high-risk level. If the temperature of the central layer begins to decrease in subsequent windows, while the temperature of the boundary layer continues to rise, it is determined that the heat from the deep layers in the previous period is being transferred outward, and the outward heat transfer status level is marked as normal transfer.

[0124] After obtaining the dynamic internal thermal state sequence, the controller determines the pulse control sequence based on the current temperature value of the central layer, the temperature change rate of the central layer, the heat accumulation level of the inner transition layer, the temperature change rate of the boundary layer, the global thermal state level, and the risk indicator of the central layer. If the current global thermal state level is in a warning state and the temperature change rate of the central layer reaches 3.1℃ / s, the controller switches the operating state from the normal operating state to the pre-control state. In the pre-control state, the controller does not reduce the target peak value of a single pulse, but compresses the conduction period of the high-heat input pulse to 15ms and sets the interval between adjacent high-heat input pulses to 25ms to reduce the accumulation of continuous heat input. If the current system has bidirectional power scheduling margin, and the central layer continues to heat up and the heat accumulation level of the inner transition layer reaches the warning level, the controller prioritizes the discharge pulse. Following the high-heat input charging pulse, the controller utilizes the slow-release characteristics of the discharge phase to reduce subsequent deep-layer heat accumulation. If the global thermal state level further escalates to a high-risk state during subsequent operation, or the central layer risk flag is activated, the controller switches to enhanced heat dissipation mode, further compressing the conduction period to 10ms and expanding the interval period to 80ms. Simultaneously, the external heat dissipation intervention level is switched from standby to pre-start or execution. If the central layer temperature change rate exceeds 2℃ / s for three consecutive update times, and the boundary layer temperature change rate is less than 0.5℃ / s, it indicates that the deep-layer temperature rise is faster than the boundary layer temperature rise, and the controller can issue an external heat dissipation pre-start request in advance. If the current central layer temperature reaches 58℃, the external heat dissipation intervention level is switched to execution, and a heat dissipation start request is output simultaneously with the subsequent pulse timing adjustment.

[0125] During the control process, the controller adjusts the timing of the charging and discharging pulses according to the pulse control sequence table, and performs graded power limiting control based on whether the internal thermal state exceeds the limit. Assuming that under enhanced heat dissipation, external heat dissipation has been initiated, but the current temperature of the central layer continues to rise to 63°C, and the rate of change of the central layer temperature is still positive, the controller enters the first power limiting level and shortens the upper limit of the cumulative conduction time to 70% of the current upper limit of the cumulative conduction time, while compressing the average duty cycle to 35%. If the central layer temperature further rises to 70°C, and the boundary layer temperature rise is less than 1°C within 5 consecutive control cycles, it is determined that the boundary layer heat dissipation has not effectively followed, and the controller enters the second power limiting level, limiting the single target peak hold time to 8ms and further narrowing the average duty cycle to 20%. If the thermal runaway precursor flag is activated but the disconnection condition has not been met, the controller enters the third power limiting level, compressing the average duty cycle to 10% and shrinking the single conduction period to the range of 5ms to 10ms to maintain the minimum continuous operation capability.

[0126] If the current temperature of the central layer further reaches the disconnection threshold, for example, when the upper limit of the safe temperature of the central layer cannot be directly obtained, the disconnection threshold is set to 80℃, and the temperature change rate of the central layer is still greater than 0 for 5 consecutive update periods, then disconnection protection is triggered; or, if the terminal voltage drops by more than 30% of the rated voltage within 2 consecutive control cycles, and the risk flag of the central layer is activated, disconnection protection is also triggered; after disconnection protection is executed, subsequent charging pulses and discharging pulses are prohibited from being released, and the control state is locked to the protection state; if an abnormal terminal voltage or a continued rise in the central layer temperature is detected after disconnection protection is executed, the controller will upgrade the protection state to the locked protection state, and automatic recovery will no longer be allowed, and manual reset will be allowed only after the fault is investigated.

[0127] During the recovery control process, if the current temperature of the central layer remains below 40℃ for more than 10 minutes and the absolute value of the temperature change rate of the central layer does not exceed 0.3℃ / s, and the abnormal operation status flag is released, then the automatic recovery conditions are met. After the automatic recovery conditions are met, the controller does not directly return to the normal operation state, but first returns to the first power limit level, and then gradually retreats to the pre-regulation state or the normal operation state according to subsequent thermal state changes. If it is currently in the locked protection state, before manual reset, it checks whether the terminal voltage abnormality flag is released, whether the central layer temperature is lower than the upper limit of the recovery temperature, whether the central layer temperature change rate has fallen back to the stable range, whether the abnormal operation status flag is released, and whether the external heat dissipation execution status is normal. After manual reset, it also first enters the first power limit level, and then gradually retreats according to thermal state changes.

[0128] Throughout the operation, the controller also performs feedback adjustments. If, after a pulse timing adjustment, the subsequent internal thermal status update results show that the rate of change of the central layer temperature has not decreased within three consecutive control cycles, it is determined that the current combination of the conduction period and the interval period is insufficient to suppress heat accumulation, and the controller will automatically request to extend the interval period or increase the level of external heat dissipation intervention in the next control cycle. If, after implementing graded power limiting control, the rate of change of the boundary layer temperature increases while the central layer temperature decreases, it is determined that deep heat is being released outward, and the current power limiting level is maintained until the release process is completed. If the voltage at the back end of the protection disconnection is still abnormal or the temperature rise continues to rise, the controller will continue to maintain the locked protection state and will not enter the automatic recovery process.

[0129] After control execution is completed, the controller generates an execution status record and operational status feedback information. The execution status record includes at least the following fields: whether the timing adjustment for the current control cycle has been completed, the current power limit level, whether disconnection protection has been triggered, and whether the lockout protection state has been entered. The operational status feedback information includes at least the following fields: timestamp, reason for state transition, core layer temperature, boundary layer temperature, current power limit level, external heat dissipation level, and recovery status. These records can be used for subsequent verification, parameter correction, and the next round of internal thermal status updates and pulse control sequence recalculation.

[0130] The above control method does not rely on a single temperature limit exceeding the limit before uniformly derating. Instead, it forms a closed-loop control link around the deep thermal state of the core, from data acquisition, internal thermal state modeling, thermal state updating, pulse sequence generation to protection execution. This allows for early intervention and regulation before the deep hot spot expands to a significant temperature rise in the shell, while also taking into account the requirements for temperature rise suppression and transient power support.

[0131] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0132] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented in software, the above embodiments can be implemented in whole or in part by a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions of the embodiments of this application are implemented in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted wirelessly or wiredly from one website, computer, server, or data center to another website, computer, server, or data center. Wired methods include optical fiber, twisted pair, coaxial cable, etc. Wireless methods include infrared, microwave, etc. Available media include any available media that can be accessed by a computer or data storage devices such as servers and data centers that contain one or more sets of available media. Available media can be magnetic media (floppy disks, hard disks, magnetic tapes), optical media (DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0133] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0134] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing temperature rise suppression in supercapacitors, characterized in that, include: S1. Collect the terminal voltage, bidirectional current and shell temperature of the supercapacitor during operation, and organize them according to a unified time reference; S2. Based on the core structure parameters, radial thermal resistance distribution, and shell boundary conditions, establish the internal thermal state distribution relationship from the core center to the shell. S3. Based on the sorted terminal voltage, bidirectional current and shell temperature, combined with the internal thermal state distribution relationship, continuously update the internal thermal state of the core at different radial positions. S4. Based on the updated internal thermal state, determine the conduction period, interval period, polarity alternation sequence, and external heat dissipation intervention sequence of the charging and discharging pulses; S5. Adjust the timing of the charging and discharging pulses according to the determined sequence, and perform graded power limiting control and disconnection protection when the internal thermal state exceeds the limit or the operating state is abnormal.

2. The method for optimizing temperature rise suppression in supercapacitors according to claim 1, characterized in that, S1 includes: The sampling points are the positive terminal, negative terminal, and shell surface temperature of the target monitoring unit, and the time is marked by a unified clock of the controller. Using high-frequency sampling time slices of voltage and current as a unified time slice, the most recent valid temperature value is used to maintain the temperature value that has not been updated. The sampled data is organized into standardized operational data frames that include timestamps, terminal voltages, bidirectional currents, casing temperatures, data validity flags, and sampling status. Verify and mark anomalies such as over-range voltage at the terminal, abnormal rate of change of bidirectional current, and sudden temperature change of the casing. Writes are made to a circular buffer using overwrite and read-write isolation methods.

3. The method for optimizing temperature rise suppression in supercapacitors according to claim 1, characterized in that, S2 includes: Based on the inner and outer dimensions of the core, core height, electrode thickness, diaphragm thickness, number of winding layers, and shell interface state of the target monitoring unit, the heat conduction path from the core center to the inner wall of the shell is discretized in layers to determine the total equivalent heat capacity and total equivalent radial thermal resistance, and the node positions, node thicknesses, node heat capacities, and inter-node thermal resistances of each layer are allocated. The basic structure is mapped and the boundary state is corrected by combining the shell temperature, ambient temperature, heat dissipation medium state, cooling level and installation posture. Generate a dataset that includes heat transfer boundary level, boundary correction status, and current update time.

4. The method for optimizing temperature rise suppression in supercapacitors according to claim 3, characterized in that, The basic structure is mapped and boundary state is corrected by combining the shell temperature, ambient temperature, heat dissipation medium state, cooling level, and installation posture, including: Based on the natural air cooling boundary, forced air cooling boundary, or liquid cooling boundary corresponding to the operating environment, and combined with the shell temperature, ambient temperature, operating status of the heat dissipation medium, cooling level, and device installation direction, the heat exchange boundary level and installation posture correction status are adjusted. When the casing temperature remains higher than the ambient temperature, adjust the boundary heat dissipation capacity to the limited level; When external cooling is activated and the casing temperature drops within a preset time window, the boundary heat dissipation capability is adjusted to the enhanced level.

5. The method for optimizing temperature rise suppression in supercapacitors according to claim 1, characterized in that, S3 includes: The running data window and the previous state of each radial level are retrieved according to a fixed update time; Heat load is determined based on the effective time ratio of bidirectional current and the amplitude of terminal voltage fluctuation. The heat input weights and heat transfer weights of the central layer, inner transition layer, outer transition layer, and boundary layer are adjusted in a layered manner; Combine the stratified temperature change rate threshold, state acceptance rules, invalid data conservative update rules, temperature probe failure handling rules, and high and low temperature anomaly handling rules; Update the current temperature value, heat accumulation level and outward heat transfer status level of each layer, and generate a dynamic internal thermal state sequence that includes the risk indicator of the central layer, the precursor indicator of thermal runaway and the global thermal state level.

6. The method for optimizing temperature rise suppression in supercapacitors according to claim 5, characterized in that, Heat load is determined based on the effective time ratio of bidirectional current and the amplitude of terminal voltage fluctuation, including: Within the preset heat load discrimination window, the effective time percentage of bidirectional current exceeding the rated continuous current discrimination threshold is statistically analyzed, and combined with the fluctuation range of terminal voltage relative to rated voltage, the heat input status of the current pulse group is classified and judged. When the effective time percentage reaches the medium sustained heat input criterion, the heat accumulation level of the central layer is increased; When the effective time percentage reaches the strong continuous heat input criterion and the terminal voltage fluctuation enters the high pulse range, the heat accumulation level of the central layer and the inner transition layer is increased.

7. The method for optimizing temperature rise suppression in supercapacitors according to claim 1, characterized in that, S4 includes: Based on the temperature status of the central layer, the heat accumulation status of the inner transition layer, the temperature rise status of the boundary layer, the global thermal status level, the risk indicators of the central layer, and the signs of thermal runaway, a state machine is constructed that includes the normal operation status, the pre-regulation status, the enhanced heat dissipation status, and the protection switching preparation status. The pulse conduction period, interval period, polarity arrangement sequence, and external heat dissipation intervention level are adjusted in stages. Based on the state entry conditions, external heat dissipation level switching conditions, protection preconditions, state rollback conditions, computing power limited branches, and field missing branches, a pulse control sequence table containing the current running state and rollback lock flags is generated.

8. The method for optimizing temperature rise suppression in supercapacitors according to claim 7, characterized in that, The pulse conduction period, interval period, polarity arrangement sequence, and external heat dissipation intervention level are adjusted in stages, including: Based on the current operating status, the conduction period of a single high-heat input pulse and the interval between adjacent pulses are adjusted in stages. When there is room for bidirectional power scheduling and the central layer continues to heat up and the heat accumulation level of the inner transition layer increases, adjust the order of charging pulses and discharging pulses. When conditions for polarity alternation are not met, extend the interval period; Based on the temperature rise rate of the central layer, the temperature rise status of the boundary layer, the temperature range of the central layer, and the risk indicators, the external heat dissipation intervention level is switched between standby, pre-start, and execution levels.

9. The method for optimizing temperature rise suppression in supercapacitors according to claim 1, characterized in that, S5 includes: Based on the current operating status and thermal risk level, a three-level power limit constraint is applied to the pulse release boundary, adjusting the cumulative conduction time, single conduction holding time, and average duty cycle respectively; When the temperature exceeds the limit, the terminal voltage drops abnormally, or the operation is abnormal, switch to the disconnect protection state or the lock protection state. Based on automatic recovery conditions or manual reset conditions, the status is rolled back, and combined with feedback adjustment rules, processing delay constraints and protection path priorities, execution status records and running status feedback information are generated.

10. The method for optimizing temperature rise suppression in a supercapacitor according to claim 9, characterized in that, When the temperature exceeds the limit, the terminal voltage drops abnormally, or there is an operational malfunction, the system switches to either the protection disconnection state or the protection lockout state, including: When the current temperature of the central layer reaches the disconnection threshold and the temperature change rate remains positive, it enters the disconnection protection state. When the terminal voltage drops abnormally within a continuous control cycle and is accompanied by the activation of the central layer risk flag, it enters the disconnection protection state. When the abnormal operation status flag indicates severe sampling distortion, control command mismatch, or external heat dissipation failure, and the global thermal status level is in protection mode, it enters the disconnect protection state. If abnormal terminal voltage or continued rise in center layer temperature persists after the protection is disconnected, the system will switch to locked protection mode.

Citation Information

Patent Citations

  • A method for charging and discharging a supercapacitor and a device for charging and discharging a supercapacitor.

    CN111628546B

  • An Online Method for Estimating the Internal Temperature of a Single Supercapacitor Cell

    CN112926186B