Safe deactivation segmented disassembly and disassembly method and system for retired power battery pack
By acquiring the acoustic impedance distribution matrix and thermal imaging path on the surface of the battery pack, and combining it with the fusion model of DS evidence theory, the problems of inaccurate defect location and geometric mismatch in the segmented disassembly and disassembly of retired power battery packs were solved, and a safe and efficient disassembly process was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-12
- Publication Date
- 2026-04-17
AI Technical Summary
Existing segmented disassembly and dismantling technology for retired power battery packs cannot accurately locate defects, leading to misjudgment of high-risk areas and geometric mismatch, resulting in decreased disassembly efficiency and increased safety risks.
By acquiring the acoustic impedance distribution matrix of the battery pack surface, a non-uniform grid thermal imaging path is generated. Combined with the fusion model of DS evidence theory, three-level risk areas are divided, and an adaptive tool path sequence is constructed to achieve accurate identification and safe disassembly of defective areas.
It achieves accurate initial screening and adaptive scanning of defective areas in battery packs, reduces the false judgment rate of high-risk areas, avoids geometric mismatch problems caused by cell bulging and deformation and connector corrosion, and improves the safety and efficiency of the disassembly process.
Smart Images
Figure CN121885829A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of retired power battery pack recycling technology, and more specifically, this application relates to a method and system for the safe deactivation and segmented disassembly of retired power battery packs. Background Technology
[0002] With the explosive growth of the new energy vehicle industry, the safe and efficient recycling of retired power battery packs has become a core part of the sustainable energy strategy. Currently, mainstream dismantling systems generally adopt a pre-set templated recommendation mechanism: a process knowledge base is built based on historical dismantling data of battery pack models, and a standardized dismantling sequence is generated through multi-dimensional feature matching (such as cell array topology, connector type, and SOC decay curve); first, the battery pack identification code is parsed, and preset force control parameters and path topology templates are called. The risk areas are marked by combining the initial screening results of infrared thermal imaging, and finally, a step-by-step dismantling instruction set is output. This recommendation mode can maintain basic operational efficiency under ideal working conditions. Its core logic relies on the geometric stability of the battery structure and the linear predictability of material aging parameters. However, in the face of retired batteries with high cycle counts, the nonlinear phase transitions and interface failures of materials caused by electrochemical decay are continuously pushing the technical boundaries of such static recommendation models.
[0003] Current technology for disassembling and reassembling retired power battery packs suffers from multiple defects: First, the aging of electrochemical materials causes dynamic drift in acoustic impedance and degradation of the thermal conduction interface, making it impossible for traditional uniform scanning paths and static compensation models to accurately locate defects. Second, the fusion of single-mode detection data and static risk models (which do not adapt to the nonlinear characteristics of battery degradation) leads to misjudgment of high-risk areas. Finally, the pre-set disassembly path template suffers from geometric mismatch (such as tool jamming) due to cell bulging and deformation / electrochemical corrosion of connectors, and lacks a real-time dynamic path optimization mechanism driven by mechanical feedback, requiring manual intervention and resulting in decreased disassembly efficiency. Therefore, a safe deactivation segmented disassembly and reassembly method and system for retired power battery packs is proposed to solve this problem. Summary of the Invention
[0004] To address the aforementioned technical problems, this technical solution provides a method and system for the safe deactivation and segmented disassembly of retired power battery packs. This solution resolves the issues raised in the background section.
[0005] To achieve the above objectives, the technical solution of the present invention is as follows:
[0006] Firstly, this application provides a method for the safe deactivation and segmented disassembly of a retired power battery pack, the method comprising:
[0007] Obtain the acoustic impedance distribution matrix on the surface of the battery pack, extract the coordinate set of the acoustic impedance abrupt change region, and generate a non-uniform grid thermal imaging path based on the spatial density distribution of the coordinate set.
[0008] Obtain the temperature distribution matrix generated by scanning along the thermal imaging path, and align it with the acoustic impedance distribution matrix using a non-uniform grid.
[0009] Based on the local acoustic impedance gradient of the acoustic impedance distribution matrix, a nonlinear gradient compensation amount is generated by combining the compensation coefficient preset by the battery pack type and the baseline offset, and the corresponding coordinate position in the temperature distribution matrix is compensated.
[0010] The local temperature gradient of the compensated temperature distribution matrix and the local acoustic impedance gradient of the acoustic impedance distribution matrix are mapped to the [0,1] interval respectively to generate thermal feature matrix and acoustic feature matrix. The two are then input into the fusion model based on DS evidence theory to output the defect risk probability, forming a risk probability matrix.
[0011] Based on the numerical range of the risk probability matrix, three levels of risk regions are divided. Based on the boundary coordinate set of these three levels of risk regions, a tool path sequence under the preset operation type constraint is constructed.
[0012] Secondly, this application provides a system for the safe deactivation and segmented disassembly of retired power battery packs, used to implement the method for the safe deactivation and segmented disassembly of retired power battery packs described in any of the above claims, including:
[0013] The acoustic conduction thermal scanning path module is used to obtain the acoustic impedance distribution matrix on the surface of the battery pack, extract the coordinate set of the acoustic impedance abrupt change region, and generate a non-uniform grid thermal imaging path based on the spatial density distribution of the coordinate set.
[0014] The grid registration module is used to acquire the temperature distribution matrix generated by scanning along the thermal imaging path and align it with the acoustic impedance distribution matrix using a non-uniform grid.
[0015] The acoustic-thermal gradient compensation module is used to generate a nonlinear gradient compensation amount based on the local acoustic impedance gradient of the acoustic impedance distribution matrix, combined with the compensation coefficient preset by the battery pack type and the baseline offset, to compensate the corresponding coordinate position in the temperature distribution matrix.
[0016] The dual-modal fusion modeling module is used to map the local temperature gradient of the compensated temperature distribution matrix and the local acoustic impedance gradient of the acoustic impedance distribution matrix to the [0,1] interval respectively, generate thermal feature matrix and acoustic feature matrix, and input the two into the fusion model based on DS evidence theory to output the defect risk probability, forming a risk probability matrix.
[0017] The risk path planning module is used to divide the risk region into three levels based on the numerical range of the risk probability matrix. Based on the boundary coordinate set of the three levels of risk region, the module constructs the tool path sequence under the preset operation type constraint. The acoustic impedance distribution matrix of the battery pack surface is used to extract the coordinate set of the acoustic impedance change region and generate a non-uniform grid thermal imaging path based on the spatial density distribution of the coordinate set.
[0018] The grid registration module is used to acquire the temperature distribution matrix generated by scanning along the thermal imaging path and align it with the acoustic impedance distribution matrix using a non-uniform grid.
[0019] The acoustic-thermal gradient compensation module is used to generate a nonlinear gradient compensation amount based on the local acoustic impedance gradient of the acoustic impedance distribution matrix, combined with the compensation coefficient preset by the battery pack type and the baseline offset, to compensate the corresponding coordinate position in the temperature distribution matrix.
[0020] The dual-modal fusion modeling module is used to map the local temperature gradient of the compensated temperature distribution matrix and the local acoustic impedance gradient of the acoustic impedance distribution matrix to the [0,1] interval respectively, generate thermal feature matrix and acoustic feature matrix, and input the two into the fusion model based on DS evidence theory to output the defect risk probability, forming a risk probability matrix.
[0021] The risk path planning module is used to divide the risk region into three levels based on the numerical range of the risk probability matrix, and to construct a tool path sequence under the preset operation type constraints based on the boundary coordinate set of the three levels of risk region.
[0022] Thirdly, this application provides a computer device, the computer device including a memory and a processor, the memory storing code, the processor being configured to acquire the code and execute the above-described method for safe deactivation and segmented disassembly of a retired power battery pack.
[0023] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for the safe deactivation and segmented disassembly of a retired power battery pack.
[0024] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0025] This application overcomes the acoustic impedance drift caused by electrochemical aging and the static compensation failure caused by the degradation of the thermal conduction interface by using dynamic acoustic impedance reference value and non-uniform grid path design, and realizes accurate initial screening and adaptive scanning of battery pack defect areas.
[0026] This application solves the defects of insufficient sensitivity of single-mode data and failure of static risk models to adapt to attenuation nonlinear characteristics by combining acoustic-thermal dual-mode nonlinear compensation with dynamic fusion of DS evidence, and realizes the quantification of defect risk probability under multi-physics coupling.
[0027] This application solves the problems of path geometry mismatch caused by cell swelling and connector corrosion, as well as the lack of dynamic optimization of pre-set templates, by using three-level risk label mapping and real-time feedback path replanning, and realizes autonomous optimization of disassembly path under safety constraints. Attached Figure Description
[0028] The disclosure of this invention is illustrated with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. Wherein:
[0029] Figure 1 The flowchart shows the safe deactivation and segmented disassembly method for retired power battery packs proposed in this invention.
[0030] Figure 2 This is a structural block diagram of the safe deactivation segmented disassembly and dismantling system for retired power battery packs proposed in this invention. Detailed Implementation
[0031] It is readily understood that, based on the technical solution of this invention, those skilled in the art can propose various interchangeable structural methods and implementations without altering the essential spirit of the invention. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solution of this invention.
[0032] During the disassembly of retired power battery packs, the dynamic drift of acoustic impedance caused by the aging of electrochemical materials and the degradation effect of the thermal conduction interface lead to a decrease in the accuracy of defect location. In particular, the uniform scanning path cannot adapt to the geometric changes caused by the nonlinear phase transition of materials; the fusion model of single-mode detection data and static risk is misjudged in high-risk areas due to its failure to adapt to the nonlinear characteristics of battery degradation. In particular, the mismatch between acoustic and thermal data causes the risk assessment to deviate from the actual state; the pre-set disassembly path template is geometrically mismatched due to the bulging deformation of the battery cells and the electrochemical corrosion of the connectors. Tool jamming requires manual intervention, which affects disassembly efficiency and increases safety risks.
[0033] For example, when disassembling a high-cycle-count retired ternary lithium battery pack, the preset force control parameters and path topology template are called. The infrared thermal imaging initial screening uses a uniform scanning path to cover the surface area. Due to the swelling and deformation of the battery cells, the actual geometric structure is offset. The acoustic impedance distribution obtained by acoustic detection is mismatched with the thermal imaging data in the coordinate system. High-risk areas are misjudged as safe areas. Electrochemical corrosion of the connectors causes the disassembly tool to get stuck in the preset path. The operation process is interrupted and manual adjustment is required, which prolongs the disassembly time and increases the possibility of electrolyte leakage.
[0034] Reference Figure 1 As shown, this application proposes a method for the safe deactivation and segmented disassembly of a retired power battery pack, including:
[0035] Obtain the acoustic impedance distribution matrix on the surface of the battery pack, extract the coordinate set of the acoustic impedance abrupt change region, and generate a non-uniform grid thermal imaging path based on the spatial density distribution of the coordinate set.
[0036] Obtain the temperature distribution matrix generated by scanning along the thermal imaging path, and align it with the acoustic impedance distribution matrix using a non-uniform grid.
[0037] It should be noted that the acoustic impedance of the battery pack surface is scanned by an ultrasonic probe to generate an acoustic impedance distribution matrix, and the temperature distribution matrix is obtained by scanning along the thermal imaging path with a thermal imager.
[0038] Based on the local acoustic impedance gradient of the acoustic impedance distribution matrix, a nonlinear gradient compensation amount is generated by combining the compensation coefficient preset by the battery pack type and the baseline offset, and the corresponding coordinate position in the temperature distribution matrix is compensated.
[0039] The local temperature gradient of the compensated temperature distribution matrix and the local acoustic impedance gradient of the acoustic impedance distribution matrix are mapped to the [0,1] interval respectively to generate thermal feature matrix and acoustic feature matrix. The two are then input into the fusion model based on DS evidence theory to output the defect risk probability, forming a risk probability matrix.
[0040] Based on the numerical range of the risk probability matrix, three levels of risk regions are divided. Based on the boundary coordinate set of these three levels of risk regions, a tool path sequence under the preset operation type constraint is constructed.
[0041] For example, a three-tiered risk area can be divided into high-risk, medium-risk, and low-risk areas, with corresponding numerical ranges of... , , ,in, This represents the probability of defect risk.
[0042] Through the above technical solution, this application achieves accurate defect identification and safe disassembly path planning by dynamically adapting to the nonlinear characteristics of battery aging. It effectively solves the problem of inaccurate defect location caused by dynamic drift of acoustic impedance and degradation of thermal conduction interface due to aging of electrochemical materials, reduces misjudgment of high-risk areas caused by single-mode detection data and static risk fusion model, and avoids geometric mismatch problems caused by cell bulging deformation and electrochemical corrosion of connectors in the pre-set disassembly path template, thus ensuring the safety and efficiency of the disassembly process.
[0043] In an optional embodiment, the coordinate set of the regions where acoustic impedance changes abruptly is extracted, specifically including:
[0044] Based on the acoustic impedance distribution matrix, calculate the absolute difference in acoustic impedance between adjacent row and column matrix elements;
[0045] Obtain the acoustic impedance reference value and its historical charge-discharge cycle count for this battery pack type. The product of the attenuation exponent raised to the power of the historical charge-discharge cycle count and the acoustic impedance reference value is used as the dynamic acoustic impedance reference value. The acoustic impedance reference value and its historical charge-discharge cycle count are stored in the battery type system library.
[0046] For example, the acoustic impedance reference value of a ternary lithium battery can be: The acoustic impedance reference value for lithium iron phosphate batteries can be: ;
[0047] It should be noted that the acoustic impedance attenuation caused by electrolyte drying and electrode pulverization follows an exponential law. Based on fitting experiments using 500 sets of disassembled retired batteries, an attenuation coefficient of 0.0015 can cover 90% of aging scenarios; therefore, the dynamic reference value for acoustic impedance is... N represents the historical number of charge-discharge cycles;
[0048] If the absolute difference of acoustic impedance is greater than the first preset ratio of the dynamic reference value of acoustic impedance, then the corresponding matrix element is recorded as a mutation point.
[0049] Merge abrupt change points whose spatial coordinate distance is less than a preset distance threshold to generate a coordinate set of acoustic impedance abrupt change regions;
[0050] It should be noted that the first preset ratio can be 30%; the minimum structural unit size of the battery module is 5mm (such as nickel sheet connectors). Abrupt points smaller than this size are considered as the same defect area to avoid over-segmentation. Therefore, the preset distance threshold can be 5mm (Euclidean distance).
[0051] Specifically, by calculating the absolute difference between adjacent row and column elements of the acoustic impedance distribution matrix, the local acoustic impedance change rate is quantified to identify potential abrupt change locations. A dynamic benchmark value is generated by constructing an attenuation exponential function with the historical charge-discharge cycle count as the power, so that the threshold automatically widens as aging intensifies. The absolute difference in acoustic impedance is compared proportionally with the dynamic benchmark value, and matrix elements that meet the conditions are marked as abrupt change points. Based on a preset distance threshold, abrupt change points are spatially clustered, and neighboring points are merged to form a continuous region coordinate set. This embodiment integrates the battery aging history into the abrupt change detection process through a dynamic benchmark value mechanism, enabling the abrupt change region extraction to adapt to different aging states and effectively solving the problem of fixed threshold failure caused by dynamic drift of acoustic impedance.
[0052] In an optional embodiment, a nonlinear gradient compensation amount is generated based on the local acoustic impedance gradient of the acoustic impedance distribution matrix, combined with a compensation coefficient preset for the battery pack type and a baseline offset. Specifically, this includes:
[0053] Taking a grid point in any non-uniform grid as the center, calculate the absolute difference of acoustic impedance of eight neighboring grids within a preset distance threshold, and take the largest absolute value as the local acoustic impedance gradient.
[0054] If the local acoustic impedance gradient is greater than the second preset ratio of the dynamic acoustic impedance reference value, the coordinate position of the grid point in the temperature distribution matrix is recorded as the point to be compensated; for example, when the acoustic impedance gradient exceeds 50% of the average interface gradient of the aged battery, the thermal imaging data will be significantly distorted, so the second preset ratio can be 47%;
[0055] Extract the compensation coefficient and baseline offset corresponding to the battery pack type, combine them with the local acoustic impedance gradient, construct a nonlinear compensation formula, and calculate the nonlinear gradient compensation amount.
[0056] The nonlinear compensation formula is as follows:
[0057] ;
[0058] In the formula, This is the nonlinear gradient compensation amount. For compensation coefficient, This is the baseline offset. For local acoustic impedance gradient, This is the gradient saturation threshold;
[0059] It should be noted that, based on the differences in thermal conductivity and damage sensitivity of different types of battery packs, the saturation critical point of the acoustic impedance gradient with respect to temperature compensation was experimentally calibrated to ensure that when... The nonlinear gradient compensation tends to stabilize, avoiding overcompensation that could lead to misjudgment; the high nickel content in the ternary lithium battery pack leads to significant thermal runaway sensitivity, requiring amplification of the acoustic impedance gradient's temperature compensation effect; the olivine structure in the lithium iron phosphate battery pack inhibits thermal diffusion, requiring a reduction in compensation intensity; the baseline offset is based on the offset value of environmental and aging interference under the calibration experimental environment.
[0060] Furthermore, the acoustic impedance gradient reflects internal structural damage to the battery pack (such as microcracks or electrolyte leakage), but the impact of this damage on temperature distribution has a physical upper limit. For example, once the electrolyte has completely leaked, no further thermal conduction hysteresis is added. The function is in Approximately linear growth (initial damage in sensitive response), when It converges quickly to the gradient saturation threshold, which matches the diminishing marginal effect of battery defects well.
[0061] For example, for a ternary lithium battery pack: the compensation coefficient, baseline offset, and gradient saturation threshold can be taken as follows: , , For lithium iron phosphate battery packs: the compensation coefficient, baseline offset, and gradient saturation threshold can be taken as follows: , , ;
[0062] Specifically, by calculating the local acoustic impedance gradient of each grid point in the non-uniform grid, the region with significant acoustic impedance change is identified as the point to be compensated; based on the compensation parameters preset by the battery pack type, the local acoustic impedance gradient is mapped into a compensation amount using a nonlinear function to correct the temperature distribution; this embodiment ensures that the temperature distribution data can truly reflect potential defects by compensating for the nonlinear change in acoustic impedance caused by battery aging.
[0063] Through the above technical solution, this application can correct the temperature distribution distortion caused by battery aging, reduce compensation error, and improve the accuracy of defect risk probability calculation.
[0064] In an optional embodiment, generating a thermal feature matrix and an acoustic feature matrix specifically includes:
[0065] Extract each grid point in the compensated temperature distribution matrix and calculate the maximum temperature difference between it and its eight neighboring grids within a preset distance threshold, which is used as the local temperature gradient reference value.
[0066] The local temperature gradient reference value is converted into a thermal feature probability value in the range [0,1] by a preset temperature gradient-probability mapping table to form a thermal feature matrix; wherein, the temperature gradient-probability mapping table is stored in the battery type database, and different nonlinear mapping curves are configured for lithium iron phosphate batteries or ternary lithium batteries.
[0067] For example, the temperature gradient-probability mapping table can be:
[0068] When the local temperature gradient At that time, the range of the thermal characteristic probability value is 0.0-0.2;
[0069] When the local temperature gradient At that time, the range of the thermal characteristic probability value is 0.2-0.5;
[0070] When the local temperature gradient At that time, the range of the thermal characteristic probability value is 0.5-0.8;
[0071] When the local temperature gradient At that time, the range of the thermal characteristic probability value is 0.8-1.0;
[0072] Calculate the absolute deviation between the local acoustic impedance gradient and the dynamic reference value of acoustic impedance, divide it by the maximum absolute deviation value in the acoustic impedance distribution matrix, generate the acoustic feature probability value in the interval [0,1], and form the acoustic feature matrix.
[0073] Specifically, the thermal anomaly boundary features are accurately identified by calculating the local temperature gradient benchmark value. This benchmark value is extracted based on the local extreme points of the compensated temperature distribution matrix, avoiding the limitations of the uniform scanning path. The local temperature gradient benchmark value is dynamically converted into a thermal feature probability value in the [0,1] interval using a temperature gradient-probability mapping table, so that its correlation with the risk probability is adaptively adjusted according to the degree of aging. At the same time, the acoustic feature probability value is generated by calculating and normalizing the absolute deviation between the local acoustic impedance gradient and the dynamic benchmark value. The dynamic benchmark value integrates historical charge and discharge cycle information, and the normalization process uses the maximum absolute deviation of the matrix as the benchmark to ensure the consistency of acoustic features in the probability space.
[0074] The above scheme can accurately reflect the nonlinear phase transition characteristics caused by battery material aging, effectively reduce the misjudgment rate of high-risk areas, and improve the reliability of defect risk probability calculation.
[0075] In an optional embodiment, the fusion model based on DS evidence theory is as follows:
[0076] ;
[0077] In the formula, For grid points The probability of defect risk, For grid points thermal feature probability value, For grid points The acoustic feature probability value, For thermal evidence weighting coefficients, For acoustic evidence weighting coefficients;
[0078] It should be noted that thermal signals are highly sensitive to internal defects in the battery pack (such as internal short circuits and lithium plating), which can directly cause local temperature rise. Actual failure data analysis shows that abnormal thermal signals account for more than 68% of defects in retired battery packs. Therefore, the thermal evidence weighting coefficient is set to 0.7 to cover 70% of the defect probability in retired battery packs. Acoustic impedance signals are highly sensitive to structural defects in the battery pack (such as cracks and desoldering) because mechanical deformation can change the propagation characteristics of sound waves. However, acoustic signals are easily affected by surface stains or assembly gaps, so their weight needs to be suppressed to avoid misjudgment. Therefore, the acoustic evidence weighting coefficient is set to 0.3.
[0079] Furthermore, the fusion model assigns higher thermal evidence weight coefficients to thermal feature probability values, enhancing the sensitivity of temperature signals to internal short-circuit defects in the battery (accounting for over 70% of failure causes); its denominator dynamically adjusts the confidence level of conflicting evidence regions, for example, when high thermal features ( ) and low acoustic characteristics ( When both thermal and acoustic features are present, the output defect risk probability decreases from 0.8 to 0.53, solving the false positive problem caused by surface contamination of the battery pack; its numerator achieves nonlinear coupling amplification of the dual-modal probability value, when both thermal and acoustic features are highly anomaly-free ( , The output risk probability reaches 0.91, which improves the detection rate of real defect risks compared with the linear weighted result;
[0080] Through the above technical solution, this application can dynamically adapt to the nonlinear characteristics of battery degradation, reduce misjudgment of high-risk areas, and improve the safety of the disassembly process of retired power battery packs.
[0081] In an optional embodiment, based on the boundary coordinate set of the three-level risk area, a tool path sequence under preset operation type constraints is constructed, specifically including:
[0082] The coordinates of the connectors in the battery pack structure information are obtained, and they are spatially associated and mapped with the boundary coordinate set of the level 3 risk area. The coordinate set of the connectors with risk level labels is output. The battery pack structure information is stored in the connector database.
[0083] Based on the connector, a preset operation type is matched, and the associated risk aversion strength coefficient is extracted;
[0084] For example, when the operation type is bolt disassembly, the risk aversion strength can be 0.3; when the operation type is laser cutting, the risk aversion strength can be 0.7; when the operation type is waterjet cutting, the risk aversion strength can be 0.9; and when the operation type is liquid nitrogen freezing, the risk aversion strength can be 0.5.
[0085] A hybrid criterion is constructed using the aforementioned risk aversion intensity coefficient and risk level label to generate the scope of the no-go zone;
[0086] The expression for the hybrid criterion is:
[0087] ;
[0088] In the formula, The radius of the restricted area. For risk aversion intensity coefficient, Risk level value;
[0089] The tool path sequence is generated with the goal of minimizing the weighted sum of the path length factor and the coverage cost factor of the path restricted area. The path length is the sum of the Euclidean distances between all adjacent path points in the tool path sequence. For example, when the three risk areas are high-risk area, medium-risk area and low-risk area respectively, the weight ratio of the path length factor and the coverage cost factor can be 1:3, 2:1 and 4:1 respectively.
[0090] The formula for calculating the coverage cost factor is as follows:
[0091] ;
[0092] In the formula, To cover the cost factor, The distance from the path point to the boundary of the restricted area;
[0093] Through the above scheme, this application can perform precise path planning based on the actual deformed connector position and the distribution of the three-level risk areas, avoiding coordinate offset problems caused by local corrosion or bulging. At the same time, it ensures that the path is adapted to different operational needs, prevents tool jamming or structural damage caused by mismatch in operational intensity, and strictly avoids high-risk areas by dynamically generating path no-go zones. Thus, it improves disassembly efficiency while ensuring safety and effectively reduces the risk of electrolyte leakage and thermal runaway.
[0094] In an optional embodiment, after generating the tool path sequence, a real-time monitoring and feedback optimization mechanism is also included:
[0095] When the average defect risk probability of three consecutive path points in the tool movement sequence exceeds a preset risk threshold, the grid where the path point is located is rescanned to update the acoustic impedance distribution matrix; the preset risk threshold can be 0.85.
[0096] Based on the updated acoustic impedance distribution matrix, the tool path sequence is regenerated starting from the grid where the current path point is located.
[0097] The above technical solution can dynamically adapt to changes in the internal state of the battery pack, effectively prevent tools from continuously entering high-risk areas, reduce the risk of electrolyte leakage or thermal runaway, and ensure the continuity and efficiency of the disassembly process.
[0098] See Figure 2 As shown, this solution proposes a safe deactivation and segmented disassembly system for retired power battery packs, used to implement the aforementioned safe deactivation and segmented disassembly method for retired power battery packs, including:
[0099] The acoustic conduction thermal scanning path module is used to obtain the acoustic impedance distribution matrix on the surface of the battery pack, extract the coordinate set of the acoustic impedance abrupt change region, and generate a non-uniform grid thermal imaging path based on the spatial density distribution of the coordinate set.
[0100] The grid registration module is used to acquire the temperature distribution matrix generated by scanning along the thermal imaging path and align it with the acoustic impedance distribution matrix using a non-uniform grid.
[0101] The acoustic-thermal gradient compensation module is used to generate a nonlinear gradient compensation amount based on the local acoustic impedance gradient of the acoustic impedance distribution matrix, combined with the compensation coefficient preset by the battery pack type and the baseline offset, to compensate the corresponding coordinate position in the temperature distribution matrix.
[0102] The dual-modal fusion modeling module is used to map the local temperature gradient of the compensated temperature distribution matrix and the local acoustic impedance gradient of the acoustic impedance distribution matrix to the [0,1] interval respectively, generate thermal feature matrix and acoustic feature matrix, and input the two into the fusion model based on DS evidence theory to output the defect risk probability, forming a risk probability matrix.
[0103] The risk path planning module is used to divide the risk region into three levels based on the numerical range of the risk probability matrix, and to construct a tool path sequence under the preset operation type constraints based on the boundary coordinate set of the three levels of risk region.
[0104] In another embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above embodiments.
[0105] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps described above.
[0106] In one embodiment, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and executes the computer instructions, causing the computer device to perform the steps described above.
[0107] The technical scope of this invention is not limited to the content described above. Those skilled in the art can make various modifications and variations to the above embodiments without departing from the technical concept of this invention, and all such modifications and variations should fall within the protection scope of this invention.
Claims
1. A method for safely deactivating and disassembling a retired power battery pack in stages, characterized in that, The method includes: Obtain the acoustic impedance distribution matrix on the surface of the battery pack, extract the coordinate set of the acoustic impedance abrupt change region, and generate a non-uniform grid thermal imaging path based on the spatial density distribution of the coordinate set. Obtain the temperature distribution matrix generated by scanning along the thermal imaging path, and align it with the acoustic impedance distribution matrix using a non-uniform grid. Based on the local acoustic impedance gradient of the acoustic impedance distribution matrix, a nonlinear gradient compensation amount is generated by combining the compensation coefficient preset by the battery pack type and the baseline offset, and the corresponding coordinate position in the temperature distribution matrix is compensated. The local temperature gradient of the compensated temperature distribution matrix and the local acoustic impedance gradient of the acoustic impedance distribution matrix are mapped to the [0,1] interval respectively to generate thermal feature matrix and acoustic feature matrix. The two are then input into the fusion model based on DS evidence theory to output the defect risk probability, forming a risk probability matrix. Based on the numerical range of the risk probability matrix, three levels of risk regions are divided. Based on the boundary coordinate set of these three levels of risk regions, a tool path sequence under the preset operation type constraint is constructed.
2. The method according to claim 1, characterized in that, The extraction of the coordinate set of the acoustic impedance abrupt change region specifically includes: Based on the acoustic impedance distribution matrix, calculate the absolute difference in acoustic impedance between adjacent row and column matrix elements; Obtain the acoustic impedance reference value and its historical charge-discharge cycle count for this battery pack type. The product of the attenuation exponent raised to the power of the historical charge-discharge cycle count and the acoustic impedance reference value is used as the dynamic acoustic impedance reference value. If the absolute difference of acoustic impedance is greater than the first preset ratio of the dynamic reference value of acoustic impedance, then the corresponding matrix element is recorded as a mutation point. Merge abrupt change points whose spatial coordinate distance is less than a preset distance threshold to generate a coordinate set of acoustic impedance change regions.
3. The method according to claim 2, characterized in that, The local acoustic impedance gradient based on the acoustic impedance distribution matrix, combined with a pre-set compensation coefficient for the battery pack type and a baseline offset, generates a nonlinear gradient compensation amount, specifically including: Taking a grid point in any non-uniform grid as the center, calculate the absolute difference of acoustic impedance of eight neighboring grids within a preset distance threshold, and take the largest absolute value as the local acoustic impedance gradient. If the local acoustic impedance gradient is greater than the second preset ratio of the acoustic impedance dynamic reference value, the coordinate position of the grid point in the temperature distribution matrix is recorded as the point to be compensated. Extract the compensation coefficient and baseline offset corresponding to the battery pack type, combine them with the local acoustic impedance gradient, construct a nonlinear compensation formula, and calculate the nonlinear gradient compensation amount. The nonlinear compensation formula is as follows: ； In the formula, This is the nonlinear gradient compensation amount. For compensation coefficient, This is the baseline offset. For local acoustic impedance gradient, This is the gradient saturation threshold.
4. The method according to claim 3, characterized in that, The generation of the thermal feature matrix and acoustic feature matrix specifically includes: Extract each grid point in the compensated temperature distribution matrix and calculate the maximum temperature difference between it and its eight neighboring grids within a preset distance threshold, which is used as the local temperature gradient reference value. By using a preset temperature gradient-probability mapping table, the local temperature gradient reference value is converted into a thermal feature probability value in the [0,1] interval to form a thermal feature matrix; Calculate the absolute deviation between the local acoustic impedance gradient and the dynamic reference value of acoustic impedance, divide it by the maximum absolute deviation value in the acoustic impedance distribution matrix, generate the acoustic feature probability value in the interval [0,1], and form the acoustic feature matrix.
5. The method according to claim 4, characterized in that, The fusion model based on DS evidence theory is as follows: ； In the formula, For grid points The probability of defect risk, For grid points thermal feature probability value, For grid points The acoustic feature probability value, For thermal evidence weighting coefficients, This represents the acoustic evidence weighting coefficient.
6. The method according to claim 1, characterized in that, The construction of a tool path sequence under preset operation type constraints based on the boundary coordinate set of the three-level risk area specifically includes: Obtain the connector coordinates from the battery pack structure information, and perform spatial association mapping between them and the boundary coordinate set of the level 3 risk area, outputting the connector coordinate set with risk level labels; Based on the connector, a preset operation type is matched, and the associated risk aversion strength coefficient is extracted; A hybrid criterion is constructed using the aforementioned risk aversion intensity coefficient and risk level label to generate the scope of the no-go zone; The tool path sequence is generated with the goal of minimizing the weighted sum of the path length factor and the coverage cost factor of the path restricted area. The path length is the sum of the Euclidean distances between all adjacent path points in the tool path sequence.
7. The method according to claim 6, characterized in that, Following the generation of the tool path sequence, a real-time monitoring and feedback optimization mechanism is also included: When the average defect risk probability of three consecutive path points in the tool movement sequence exceeds the preset risk threshold, the grid where the path point is located is rescanned to update the acoustic impedance distribution matrix. Based on the updated acoustic impedance distribution matrix, the tool path sequence is regenerated, starting from the grid where the current path point is located.
8. A safe deactivation and segmented disassembly system for retired power battery packs, characterized in that, A method for safely deactivating and disassembling a retired power battery pack in segments as described in any one of claims 1-7 includes: The acoustic conduction thermal scanning path module is used to obtain the acoustic impedance distribution matrix on the surface of the battery pack, extract the coordinate set of the acoustic impedance abrupt change region, and generate a non-uniform grid thermal imaging path based on the spatial density distribution of the coordinate set. The grid registration module is used to acquire the temperature distribution matrix generated by scanning along the thermal imaging path and align it with the acoustic impedance distribution matrix using a non-uniform grid. The acoustic-thermal gradient compensation module is used to generate a nonlinear gradient compensation amount based on the local acoustic impedance gradient of the acoustic impedance distribution matrix, combined with the compensation coefficient preset by the battery pack type and the baseline offset, to compensate the corresponding coordinate position in the temperature distribution matrix. The dual-modal fusion modeling module is used to map the local temperature gradient of the compensated temperature distribution matrix and the local acoustic impedance gradient of the acoustic impedance distribution matrix to the [0,1] interval respectively, generate thermal feature matrix and acoustic feature matrix, and input the two into the fusion model based on DS evidence theory to output the defect risk probability, forming a risk probability matrix. The risk path planning module is used to divide the risk region into three levels based on the numerical range of the risk probability matrix, and to construct a tool path sequence under the preset operation type constraints based on the boundary coordinate set of the three levels of risk region.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method described in any one of claims 1-7.
10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-7.