Environment-friendly manufacturing method of high-capacity battery

Abstract

The invention relates to the technical field of battery manufacturing, and discloses an environment-friendly manufacturing method of a high-capacity battery. The method comprises the following steps: firstly, collecting relevant data of sustainable material attributes, green process monitoring and environment footprints, and extracting an initial feature time sequence from the relevant data; a fusion feature set of material-process-environment interaction is extracted, a multi-scale model structure is constructed in combination with data acquisition precision of different making partitions, and multi-resolution feature mapping is generated; performing trend analysis on the time evolution data to generate a battery performance degradation spectrogram; stress response distribution of the core production stage is determined through simulation calculation, and a manufacturing state response data set is obtained; and identifying the number of performance degradation cycles, calculating a residual capacity value by fusing a material degradation model, and evaluating the total life of the battery produced in an environment-friendly manner after integration. According to the method, environmental protection and green consideration of the whole production process is realized through multi-dimensional data fusion and multi-scale analysis.

Description

Technical Field

[0001] This invention relates to the field of battery manufacturing technology, specifically to an environmentally friendly and green manufacturing method for high-capacity batteries. Background Technology

[0002] Currently, high-capacity batteries are increasingly widely used in new energy vehicles, energy storage systems, and other fields, and their production scale continues to expand. However, in the existing battery manufacturing process, the selection of materials largely relies on traditional chemical raw materials. These materials are not only scarce resources, but also easily generate a large amount of pollutants during the mining and processing stages, placing a burden on the ecological environment. At the same time, the production process mostly adopts a single fixed process, lacking consideration of the dynamic relationship between material properties and process parameters, resulting in high energy consumption during the production process, and some stages also have problems with excessive waste emissions. In terms of environmental impact assessment, existing methods mostly focus on pollutant detection at the end of the production process, making it difficult to track the environmental footprint throughout the entire production cycle in real time. Mismatches between material properties and process parameters often lead to unstable battery performance, with some batches experiencing excessively rapid capacity decay, wasting resources and increasing the pressure on subsequent recycling and disposal. Furthermore, the accuracy of process parameter data collection varies across different production zones, making it difficult for traditional models to integrate multi-scale data. This results in insufficient accuracy in predicting battery performance degradation and an inability to adjust production strategies in a timely manner to reduce environmental impact. Current technologies primarily assess battery life based on finished product performance testing, neglecting the cumulative impact of key production stages on overall battery life. This makes it difficult to effectively achieve the goals of environmentally friendly and green manufacturing. Furthermore, the lack of systematic analysis of the interactions between materials, processes, and the environment during production leads to low resource utilization rates, failing to meet the requirements of sustainable development. Summary of the Invention

[0003] The purpose of this invention is to provide an environmentally friendly and green manufacturing method for high-capacity batteries, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides an environmentally friendly green manufacturing method for high-capacity batteries, the method comprising: Data sets of sustainable material properties, green process monitoring data, and environmental footprint indicators related to the battery production process are collected, and initial feature time series are extracted from these data sets. The manufacturing process of the high-capacity battery includes multiple core production stages. Extract a fusion feature set of material-process-environment interaction from the initial feature time series; based on the fusion feature set and the location data acquisition accuracy of different battery manufacturing zones, construct multi-scale model structures with different granularities respectively, and output multi-resolution feature mapping based on the multi-scale model structure generation results. A trend analysis operation is performed on the time evolution data of the multi-resolution feature mapping to generate a battery performance degradation spectrum. Based on the battery performance degradation spectrum, simulation calculation technology is used to determine the stress response distribution of the multiple core production stages and obtain a battery manufacturing state response dataset. Based on the battery manufacturing status response dataset, the number of performance degradation cycles in the core production stage is identified; and by integrating a preset battery material degradation model, the remaining capacity value of the core production stage is calculated. Then, by integrating the remaining capacity values, the overall lifespan of the high-capacity battery in environmentally friendly green manufacturing is evaluated.

[0005] Preferably, the granularity of the multi-scale model structure is set according to the model unit size parameter, the total number of model elements parameter, and the model density parameter; Based on the fused feature set and combined with the location data acquisition accuracy of different battery manufacturing zones, multi-scale model structures with different granularities are constructed respectively. The multi-scale model structure generates multi-resolution feature maps, including: The battery manufacturing area is divided into several independent sub-manufacturing blocks; Create multiple parallel computing task units, each of which is associated with an independent sub-production block; Based on the location data acquisition accuracy of each independent sub-production block, the model unit size parameters, total number of model elements parameters, and model density parameters of each independent sub-production block are determined; The multiple parallel computing task units are coordinated to generate target multi-scale models for each independent sub-production block based on the model unit size parameters, total number of model elements parameters, and model density parameters of each independent sub-production block. The multi-scale model of the target is integrated with the fused feature set and each independent sub-production block to output the multi-resolution feature map.

[0006] Preferably, the coordination of the multiple parallel computing task units to generate the target multi-scale model for each independent sub-production block according to the model unit size parameter, the total number of model elements parameter, and the model density parameter of each independent sub-production block includes: The multiple parallel computing task units are controlled to generate preliminary geometric information of the model surface based on the model unit size parameters, total number of model elements parameters, and model density parameters of each independent sub-production block; The multiple parallel computing task units are controlled to construct three-dimensional solid models of each independent sub-production block and obtain model volume topology data based on the geometric information of the model surface; Based on the model topology data, extract the vertex set and volume element set of each 3D entity model; and perform a unique encoding operation on the vertex set and volume element set of each 3D entity model to obtain a unique identification identifier for each vertex set and a unique identification identifier for each volume element set. Control the data synchronization transmission between the multiple parallel computing task units, so that the volume unit sets and exclusive identification marks between adjacent independent sub-production blocks are mutually projected and mapped to generate a projection mapping dataset. Based on the projection mapping dataset, the 3D solid model of each independent sub-production block is adaptively adjusted to obtain the target multi-scale model of each independent sub-production block.

[0007] Preferably, the step of extracting the initial feature time series from the sustainable material property dataset, green process monitoring flow data, and environmental footprint indicator information includes: Standardization transformation is performed on sustainable material property datasets, green process monitoring stream data, and environmental footprint index information to output a normalized dataset; A preset feature extraction mechanism is used to extract material chemical features, process dynamic features, and environmental interaction features from a normalized dataset, and these features are then combined to form the initial feature time series.

[0008] Preferably, the extraction of the fused feature set of material-process-environment interaction from the initial feature time series includes: Extract a primary feature set that couples material properties, process parameters, and environmental factors from the initial feature time series; When the feature dimension of the primary feature group exceeds the preset dimension threshold, a dimensionality reduction operation is performed on the primary feature group to obtain a dimensionality-reduced feature dataset. Correlation analysis is performed on the time series of the dimensionality-reduced feature dataset to enhance the dynamic correlation between material property features, process parameter features, and environmental factor features, thereby generating the fused feature set.

[0009] Preferably, the high-capacity battery manufacturing process is connected to at least one green energy supply device; the determination of the stress response distribution of the multiple core production stages based on battery performance degradation spectrum using simulation calculation technology to obtain a battery manufacturing state response dataset includes: Construct a production system simulation model for the high-capacity battery manufacturing process; construct separate device simulation models for each green energy supply unit, and integrate all device simulation models into the production system simulation model to form a comprehensive simulation model; External excitation input is constructed based on the battery performance degradation spectrum; simulation calculation and analysis are performed based on the comprehensive simulation model and the external excitation input, combined with the preset operating constraints, to obtain the stress response distribution of each core production stage; Based on the stress response distribution of each core production stage, analyze the time variation curves and spatial distribution patterns of stress in each core production stage. By merging the time variation curves and spatial distribution patterns of stress in each core production stage, a dynamic stress evolution map of the high-capacity battery manufacturing process is generated, and a battery manufacturing state response dataset is derived based on the dynamic stress evolution map.

[0010] Preferably, the method further includes: Responding to optimization instructions for a specific production stage, the optimization instructions include parameters for calculating accuracy requirements; Based on the optimization instructions, feature data sequences corresponding to specific production stages are extracted from the battery manufacturing status response dataset. Based on the required accuracy parameters, the performance optimization analysis results for a specific production stage are calculated using feature data sequence analysis.

[0011] Preferably, the method further includes: The battery manufacturing facility is divided into multiple production unit areas. Records of renewable resource consumption, clean energy use data and carbon emission monitoring streams of each production unit area are obtained to form a multi-source environmental data set. Resource efficiency assessment indicators and environmental impact assessment indicators are set based on multi-source environmental protection data sets, and initial optimization potential values ​​are generated through a dual threshold triggering mechanism and a dynamic decay model. Based on the spatial correlation characteristics of adjacent production unit areas, the initial optimization potential value of each production unit area is dynamically corrected, and the corrected optimization potential value is output. The revised optimization potential values ​​are sorted in descending order to generate a priority sorting form, which is then integrated with the production management system for visual display.

[0012] Preferably, the method further includes: Obtain a set of battery manufacturing record data for multiple batches corresponding to the target production line. The set of manufacturing record data includes process description text information, process sensor time series data, and raw material sustainability indicators for each production batch. Semantic parsing is performed on the process description text information to output the process semantic feature vector. Fluctuation feature extraction is performed on the process sensor time series data to output the process dynamic feature vector. The process semantic feature vector and the process dynamic feature vector are then fused together to generate a fused production feature map. The pre-trained anomaly analysis model is invoked to perform anomaly contribution allocation on the fusion production feature map, generating anomaly correlation distribution map corresponding to the production batch. The anomaly correlation distribution map includes the distribution of anomaly correlation strength between raw material sustainability indicators and each production process. Based on the distribution of abnormal correlation strength, the joint root cause tracing calculation of the sustainability indicators of raw materials and the textual information of process description is performed to output the environmental anomaly diagnosis results of the production batch. The environmental anomaly diagnosis results are used to locate the production process and material defect type of the source of the problem. Based on the environmental anomaly diagnosis results, a production parameter optimization strategy is generated and transmitted to the production control system to trigger parameter update operations.

[0013] Preferably, the method further includes: Obtain the current battery manufacturing parameter configuration set and establish a database of battery quality and environmental indicators associated with the current battery manufacturing parameter configuration set; A set of targeted optimization objectives is configured based on a database of battery quality and environmental protection indicators. The set of targeted optimization objectives includes capacity maximization optimization, material renewability optimization, energy efficiency optimization, and pollutant emission reduction optimization. Determine the optimization priority identifiers in the set of targeted optimization objectives, wherein the optimization priority identifiers include primary optimization objectives, secondary optimization objectives, and stability objectives; After selecting the key optimization parameter set using the primary and secondary optimization objectives, an impact assessment of the targeted optimization objectives of the key optimization parameter set is performed. After establishing parameter optimization constraints based on the impact assessment results of the targeted optimization objectives, parameter-oriented optimization calculations are performed, and the results of the parameter-oriented optimization calculations are used to update the environmentally friendly manufacturing parameters for high-capacity batteries.

[0014] Compared with the prior art, the beneficial effects of the present invention are: By collecting sustainable material property datasets, green process monitoring data, and environmental footprint indicators, we can comprehensively capture multi-dimensional information about materials, processes, and the environment during the production process. The initial feature time series extracted from this information provides rich foundational data for subsequent analysis, helping to gain a deeper understanding of the dynamic changes of each factor throughout the production cycle. Extracting a fusion feature set of material-process-environment interactions from the initial feature time series can transform previously scattered, single-dimensional data into closely related comprehensive information, revealing the intrinsic connections between material properties, process parameters, and environmental impacts. The multi-scale model structure with varying granularities built upon this fusion feature set can adapt to the location data acquisition precision of different production zones, and the generated multi-resolution feature maps can reflect the production status at different scales, providing a more comprehensive understanding of the production process. Trend analysis of the time evolution data from multi-resolution feature mapping generates a battery performance degradation spectrum that visually presents the trajectory of battery performance changes with the production process, facilitating the identification of key performance degradation nodes. Simulation calculations are used to determine the stress response distribution at core production stages, and the resulting battery manufacturing state response dataset clearly demonstrates the impact of process parameters at each stage on battery state, providing direction for optimizing the production process. By identifying the number of performance degradation cycles in key production stages and integrating them with a pre-defined battery material degradation model to calculate the remaining capacity, the impact of each stage on battery life can be integrated, considering environmental protection and green attributes from the source of production. By integrating various remaining capacity values ​​to assess the overall lifespan of the battery's environmentally friendly manufacturing, environmental protection concepts can be integrated into the entire production process. This makes material selection more aligned with sustainable development needs, process adjustments more in line with environmentally friendly standards, reduces resource waste and environmental burden during production, and achieves synergistic optimization of materials, processes, and the environment. This method does not rely on single data or fixed models. Instead, it uses multi-dimensional data fusion and multi-scale analysis to make environmental considerations in the production process more systematic and dynamic. The sustainable properties of materials are incorporated into production decisions, process parameters are adjusted in real time based on material characteristics and environmental impact, and environmental footprint indicators can be monitored in real time and fed back to the production process, forming a virtuous cycle. This allows the production of high-capacity batteries to not only improve performance but also align more closely with the direction of green and environmentally friendly development. Attached Figure Description

[0015] Figure 1 This is a schematic diagram illustrating the working principle of the environmentally friendly green manufacturing method for the high-capacity battery described in this invention. Figure 2 Flowchart for constructing a multi-scale model structure; Figure 3 The flowchart for feature set extraction is shown below; Figure 4 The flowchart for stress response distribution simulation; Figure 5 A flowchart for assessing the potential for optimizing environmental data. Detailed Implementation

[0016] 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.

[0017] Please see Figure 1 This invention provides an environmentally friendly green manufacturing method for high-capacity batteries, the method comprising: The process involves collecting sustainable material property datasets, green process monitoring stream data, and environmental footprint indicators related to battery production. Initial feature time series are then extracted from these datasets. The high-capacity battery manufacturing process comprises multiple core production stages. The sustainable material property dataset covers attributes such as material origin, renewable proportion, and degradation rate. The green process monitoring stream data includes real-time operating parameters of each production device, energy consumption data, and waste emission data. The environmental footprint indicators involve quantitative indicators such as carbon emissions, water consumption, and land occupation during the production process. The extraction of the initial feature time series must retain the temporal dimension characteristics of each data point to ensure the continuity and integrity of the data along the timeline.

[0018] A fusion feature set of material-process-environment interactions is extracted from the initial feature time series. Based on this fusion feature set and the location data acquisition accuracy of different battery manufacturing zones, multi-scale model structures with different granularities are constructed. Multi-resolution feature maps are output based on the results generated from the multi-scale model structures. The extraction of the fusion feature set requires algorithmic mining of the correlation between material properties, process parameters, and environmental indicators, such as the synergistic impact of material composition changes and process temperature fluctuations on carbon emissions. The construction of the multi-scale model structure requires adjusting the level of detail of the model according to the differences in location data acquisition accuracy of different manufacturing zones to adapt to the data characteristics of different regions.

[0019] Trend analysis is performed on the time evolution data of the multi-resolution feature map to generate a battery performance degradation spectrum. The trend analysis captures the patterns of change in the multi-resolution feature map over time, identifies feature change patterns that may lead to battery performance degradation, and then visually presents the process and extent of performance degradation in the form of a spectrum.

[0020] Based on the battery performance degradation spectrum, simulation calculation technology is used to determine the stress response distribution of the multiple core production stages, thereby obtaining a battery manufacturing state response dataset. The simulation calculation technology needs to simulate the stress changes of each core production stage under different operating conditions, including mechanical stress and thermal stress, to obtain comprehensive state response data.

[0021] Based on the battery manufacturing state response dataset, the number of performance degradation cycles in the core production stage is identified. A pre-defined battery material degradation model is then integrated to calculate the remaining capacity value for the core production stage. Finally, these remaining capacity values ​​are combined to assess the overall environmentally friendly lifespan of the high-capacity battery. The identification of the performance degradation cycle count requires consideration of the periodic variation characteristics in the state response data, while the material degradation model is based on the pre-defined physicochemical properties of the materials. The integration of remaining capacity values ​​achieves a comprehensive assessment of the overall battery manufacturing lifespan.

[0022] Example 1: See Figure 2 The granularity of the multi-scale model structure is set based on the model unit size parameter, the total number of model elements parameter, and the model density parameter.

[0023] Based on the fused feature set and combined with the location data acquisition accuracy of different battery manufacturing zones, multi-scale model structures with different granularities are constructed. Multi-resolution feature maps are then generated based on the results of these multi-scale model structures. The specific process is as follows: The battery manufacturing area is divided into several independent sub-manufacturing blocks. This division requires comprehensive consideration of the battery production process flow, the spatial layout of production equipment, and the functional independence of each stage. Each sub-manufacturing block contains a relatively complete production sub-process, and the block boundaries match the physical isolation or equipment layout in actual production to ensure the relevance and accuracy of subsequent model construction.

[0024] Multiple parallel computing task units are created, each associated with an independent sub-production block. The creation of these task units relies on a distributed computing architecture, with each unit equipped with independent computing resources, including processors and memory, and data exchange between units is achieved through an internal communication protocol. The association between task units and sub-production blocks uses a one-to-one correspondence, ensuring that the model computation of each sub-production block is handled by its dedicated task unit, avoiding computing resource conflicts and data confusion.

[0025] Based on the location data acquisition accuracy of each independent sub-production block, the model unit size parameters, total number of model elements, and model density parameters for each independent sub-production block are determined. Location data acquisition accuracy is comprehensively determined by the sensor's sampling frequency, measurement error range, and data transmission rate. For areas with high acquisition accuracy, such as electrode material coating areas equipped with high-precision laser sensors, the model unit size parameter is set to a smaller value; for areas with low acquisition accuracy, such as warehouse transfer areas relying on manual inspection records, the model unit size parameter can be increased to 10mm × 10mm, the total number of model elements reduced to below 1000, and the model density parameter configured at 5 elements per cubic millimeter.

[0026] Multiple parallel computing task units are coordinated to generate target multi-scale models for each independent sub-production block based on its model unit size, total number of model elements, and model density parameters. This coordination process is implemented through a central control module, which monitors the computation progress of each task unit in real time. When a unit experiences a computational delay, it dynamically allocates idle resources from other units to provide support. Simultaneously, the central control module uniformly calibrates the time base of each task unit, ensuring that model computations for all sub-production blocks are performed within the same time dimension, avoiding model integration deviations caused by time asynchrony.

[0027] The system integrates the fusion feature set and the target multi-scale models of each independent sub-fabrication block to output a multi-resolution feature map. The integration process first standardizes the material property data, process parameter data, and environmental index data in the fusion feature set to match their data format with the model outputs of each sub-fabrication block. Based on the spatial relationships of the sub-fabrication blocks, the target multi-scale models are stitched together to form a complete battery fabrication area model, and the corresponding data from the fusion feature set are embedded into the corresponding positions in the model. The output multi-resolution feature map is presented in the form of a 3D atlas. Different sub-fabrication blocks exhibit different levels of detail resolution based on differences in model parameters. High-resolution areas clearly display subtle features of material microstructure, process parameter fluctuations, and environmental index changes, while low-resolution areas present the overall state with macroscopic trends.

[0028] Specifically, multiple parallel computing task units are coordinated to generate target multi-scale models for each independent sub-production block based on the model unit size parameters, total number of model elements, and model density parameters. This includes: Multiple parallel computing task units are controlled to generate preliminary model surface geometry information based on the model unit size parameters, total number of model elements, and model density parameters of each independent sub-fabrication block. This surface geometry information is generated using a 3D modeling algorithm, encompassing the boundary contours of the sub-fabrication blocks, equipment shapes, and material stacking patterns. For example, the model surface of the electrode preparation area needs to accurately reproduce the nozzle trajectory of the coating machine, the running path of the conveyor belt, and the cavity structure of the drying equipment. The generation of its geometry information requires combining CAD design drawings and on-site scanning data to ensure a high degree of consistency with the actual production scenario.

[0029] Multiple parallel computing task units are controlled to construct 3D solid models of independent sub-production blocks based on the surface geometry information of the model and obtain the model volume topology data. The construction of the 3D solid model is based on the surface geometry information, and the internal structural details are filled in. The model volume topology data is extracted through topology algorithms, including vertex coordinates, edge connectivity, face composition, and volume spatial distribution. This data is stored in matrix form, and each vertex, edge, face, and volume is assigned a unique identifier to facilitate subsequent data retrieval and model adjustment.

[0030] Based on the model's topology data, the vertex set and volume element set of each 3D solid model are extracted. A unique encoding operation is then performed on each vertex set and volume element set to obtain a unique identifier for each vertex set and volume element set. The vertex set contains the 3D coordinates of all vertices in the model, and the volume element set contains the spatial extent and attribute parameters of all volume elements. The unique encoding is implemented using a 16-bit binary number; the first 8 bits represent the sub-fabrication block number, and the last 8 bits represent the element's sequence number within the block, ensuring that the identifier for each vertex and volume element is unique within the battery fabrication area model.

[0031] The system controls the synchronous data transmission between multiple parallel computing task units, enabling mutual projection mapping of volume unit sets and unique identifiers between adjacent independent sub-fabrication blocks, generating a projection mapping dataset. Data synchronization is achieved via a high-speed fiber optic channel, with the transmission frequency consistent with the model computation frequency, i.e., data is transmitted once every 0.1 seconds. The projection mapping of volume unit sets in adjacent blocks employs an overlapping region matching algorithm, pairing volume units at boundaries according to their spatial location. For example, at the boundary between the electrode preparation area and the cell assembly area, the coated electrode volume units are mapped to the clamping equipment volume units at the assembly station, recording their spatial distance, contact area, and relative motion trajectory, forming a projection mapping dataset containing positional relationships, attribute associations, and dynamic interaction information.

[0032] The projection mapping dataset adaptively adjusts the 3D solid models of each independent sub-fabrication block to obtain the target multi-scale model for each sub-fabrication block. The adjustment process corrects boundary mismatches in the projection mapping dataset; when the difference in volumetric element attribute parameters exceeds a preset threshold, the material properties or process parameters of the relevant elements are recalculated. For areas with frequent dynamic interactions, such as the junctions of electrode transfers, the density of model volumetric elements is increased to more precisely capture detailed changes during the interaction process. The final target multi-scale model maintains the independent characteristics of each sub-fabrication block while ensuring seamless connections between adjacent blocks.

[0033] Example 2: See Figure 3 The initial feature time series was extracted from sustainable material property datasets, green process monitoring flow data, and environmental footprint indicator information. The specific operations are as follows: Standardization transformation was performed on sustainable material property datasets, green process monitoring stream data, and environmental footprint indicator information to output a normalized dataset. The sustainable material property dataset covers material origin information, renewable component proportions, natural degradation cycles, and heavy metal content; the green process monitoring stream data includes real-time power consumption, operating temperature, material transfer speed, and waste discharge of various equipment on the production line; and the environmental footprint indicator information involves water consumption per unit product, total carbon emissions, volatile organic compound emissions, and solid waste generation during the production process. The standardization transformation first unified the recording format of various data types, converting unstructured data from different devices into structured tables, ensuring that each data entry includes a timestamp, data type identifier, and specific value. Next, data with different units were processed, mapping data such as the renewable component proportion of materials, equipment power consumption, and water consumption per unit product to a 0-1 value range through linear transformation. Simultaneously, a sliding window method was used to identify outliers in the data; for data exceeding the normal fluctuation range, the average value of adjacent time periods was used for correction to avoid extreme values ​​interfering with subsequent analysis.

[0034] A pre-defined feature extraction mechanism is used to extract material chemical characteristics, process dynamic characteristics, and environmental interaction characteristics from a normalized dataset, merging them to form an initial feature time series. Material chemical characteristics include the chemical stability of materials at different temperatures, reactivity with other materials, and molecular structural stability; process dynamic characteristics include the temperature rise curve during equipment startup, parameter fluctuation range during stable operation, and energy consumption reduction trend during shutdown; environmental interaction characteristics include changes in carbon emissions per unit time during production, energy consumption fluctuations during waste treatment, and time-dependent changes in water resource recycling rates. The feature extraction mechanism combines a sliding time window with a feature selection algorithm. The length of each time window is set according to the duration of the production process, typically 10 minutes. Within each window, statistics on material chemical characteristics, process dynamic characteristics, and environmental interaction characteristics are calculated. These statistics are arranged chronologically to form an initial feature time series containing the changes of material, process, and environmental characteristics over time. Each data point in the series corresponds to a feature summary result for one time window.

[0035] The process of extracting a fused feature set of material-process-environment interaction from the initial feature time series is as follows: Primary feature groups, representing the coupling of material properties, process parameters, and environmental factors, are extracted from the initial feature time series. By analyzing the changing trends of various features in the initial feature time series, the correlations among the three are identified. For example, when the chemical stability of a material decreases, it is observed whether the corresponding process parameters show an increase in temperature or pressure, and whether this change leads to an increase in carbon emissions in environmental indicators. Material property features, process parameter features, and environmental factor features with significant correlations are combined to form primary feature groups. The extraction of primary feature groups needs to cover all core stages of battery production. For example, in the electrode preparation stage, the coupling relationship between the particle size distribution of positive and negative electrode materials and the rotation speed and energy consumption per unit time of the stirring equipment; in the assembly stage, the coupling relationship between the temperature resistance of the separator material and the heating temperature of the packaging equipment and the amount of exhaust gas emitted in the workshop, etc., ensures that the primary feature groups can reflect the interactions of materials, processes, and the environment in different production stages.

[0036] When the feature dimension of the primary feature group exceeds a preset dimensionality threshold, dimensionality reduction is performed on the primary feature group to obtain a dimensionality-reduced feature dataset. The preset dimensionality threshold is determined based on the computational resources for data processing and the complexity requirements of subsequent models, and is typically set to 500 dimensions. Dimensionality reduction is initiated when the number of features in the primary feature group exceeds this threshold. The dimensionality reduction process preserves the main variation information between features and removes redundant features. Dimensionality reduction does not change the time-series characteristics of the features; it only reduces the number of features, thus reducing the data volume of the processed dimensionality-reduced feature dataset while maintaining core information.

[0037] Correlation analysis is performed on the time series data of the dimensionality-reduced feature dataset to enhance the dynamic correlation between material properties, process parameters, and environmental factors, thereby generating a fused feature set. Correlation analysis is achieved by calculating the degree of synchronous change of different features over time. For example, the correlation coefficient between the degradation rate of materials and temperature changes in the process, as well as carbon emissions in the environment, is calculated to identify the lead-lag relationship among the three. For features with strong correlations, a feature fusion algorithm is used to merge them into a single comprehensive feature, which simultaneously incorporates information from material properties, process parameters, and environmental factors. For instance, the cycle life of the cathode material, the temperature characteristics of the calcination process, and the energy consumption characteristics of the corresponding stages are fused into a new feature that reflects the comprehensive impact of the material on environmental resource consumption under specific process conditions. In this way, the generated fused feature set not only contains individual information from each feature but also highlights the dynamic interaction among the three, forming a feature set that comprehensively reflects the synergistic effect of materials, processes, and the environment.

[0038] Example 3: See Figure 4The high-capacity battery manufacturing process is connected to at least one green energy supply device, including solar power generation equipment, wind power generation equipment, biomass energy conversion devices, etc., which are connected to the power supply network of the battery production line through an energy management system to achieve dynamic allocation and supply of energy.

[0039] Based on the battery performance degradation spectrum, simulation calculation technology is used to determine the stress response distribution of multiple core production stages, and obtain a battery manufacturing state response dataset. The specific process is as follows: A production system simulation model for high-capacity battery manufacturing is constructed. Separate simulation models for each green energy supply device are also constructed. All device simulation models are then integrated into the production system simulation model to form a comprehensive simulation model. The production system simulation model includes equipment models, material handling models, and process logic models for each core stage of battery production. The equipment models reflect the physical parameters and operating characteristics of equipment such as mixers, coating machines, and rolling mills. The material handling models simulate the flow paths and time points of materials such as electrodes and electrolytes between different devices. The process logic models define the sequence and connection conditions of each production stage. The device simulation models for the green energy supply devices include energy output curves, conversion efficiency parameters, and storage capacity parameters. For example, the model for solar power generation equipment needs to reflect the relationship between light intensity and power generation, and the model for wind power generation equipment needs to include the correspondence between wind speed and output power. The integration process is achieved through an energy interface module, which matches the output of the green energy supply devices with the energy consumption requirements of the production system, forming a comprehensive simulation model covering the production process and energy supply.

[0040] External excitation inputs are constructed based on battery performance degradation spectra. Simulation calculations and analyses are performed according to the comprehensive simulation model and the external excitation inputs, combined with preset operational constraints, to obtain the stress response distribution of each core production stage. The battery performance degradation spectra contain information on the time-varying characteristics of battery capacity decay rate and internal resistance growth trend. This information is converted into external excitations for the production system. For example, when battery performance degrades to a certain threshold, instructions are simulated to apply additional loads to the production equipment or adjust process parameters. Preset operational constraints include the maximum operating temperature of the equipment, minimum safe pressure, and the humidity range of the materials. The simulation calculations and analyses use numerical iteration methods to solve for the stress distribution of the production system under external excitations, involving various types of stress, including mechanical stress, thermal stress, and chemical stress.

[0041] Based on the stress response distribution at each core production stage, the temporal variation curves and spatial distribution patterns of stress at each core production stage are analyzed. The temporal variation curves record the numerical fluctuations of stress during the production process, such as the change in pressure applied by the roller press over time during the electrode rolling stage; the spatial distribution patterns show the differences in stress distribution inside the equipment and in the material, such as the pressure distribution at the coating machine nozzle and the stress gradient inside the electrode.

[0042] By merging the temporal variation curves and spatial distribution patterns of stress at each core production stage, a dynamic stress evolution map of the high-capacity battery manufacturing process is generated. Based on this dynamic stress evolution map, a battery manufacturing state response dataset is derived. The dynamic stress evolution map presents the comprehensive changes of stress in time and space in a three-dimensional visualization. The exported battery manufacturing state response dataset contains detailed data such as stress values ​​and rates of change at each time point and location.

[0043] The system responds to optimization instructions for specific production stages, which include parameters for required computational accuracy. These instructions are generated by the production monitoring system based on real-time production status. When the stress value of a production stage approaches the constraint threshold, an optimization instruction is automatically issued. The required computational accuracy parameters are set according to the importance of that production stage; critical stages have higher accuracy requirements, while non-critical stages can have lower accuracy to reduce computational resource consumption.

[0044] Based on the optimization instructions, feature data sequences corresponding to specific production stages are extracted from the battery manufacturing status response dataset. The extraction process uses the spatial coordinate information in the optimization instructions to locate the equipment or material area of ​​the specific production stage, and extracts stress-related time-series data within that area, including stress peaks, troughs, and durations.

[0045] Based on the required computational accuracy parameters, performance optimization analysis results for a specific production stage are calculated using feature data sequence analysis. During the analysis, the required computational accuracy parameters determine the level of detail in the data processing; when high accuracy is required, smaller time intervals and denser spatial sampling points are used for analysis. The following formulas are used in the analysis:

[0046] in, This indicates the target adjustment time for performance optimization analysis. This represents the initial response time in the feature data sequence. This represents the standard deviation of stress fluctuations in the characteristic data sequence. and These represent the time weighting coefficient and the stress weighting coefficient, respectively, both ranging from 0 to 1, and are determined based on the process characteristics of a specific production stage. This formula calculates the time points when production parameters need adjustment. Combined with other information from the characteristic data sequence, it generates performance optimization analysis results, including equipment parameter adjustment suggestions and process time optimization schemes.

[0047] Example 4: See Figure 5The battery manufacturing facility is divided into multiple production unit areas, based on factors including differences in production functions, the concentration of equipment layout, and the coverage of environmental monitoring points. For example, it can be divided into areas for cathode material preparation, anode material preparation, cell assembly, packaging and testing, and waste disposal. Records of renewable resource consumption, clean energy usage data, and carbon emission monitoring streams are acquired from each production unit area to form a multi-source environmental data set. Renewable resource consumption records include the daily and weekly consumption of recycled aluminum, recycled plastics, and other materials used in each area; clean energy usage data covers the hourly power supply and monthly total consumption of clean energy sources such as solar and wind power connected to each area; and carbon emission monitoring streams include real-time carbon emission concentrations and hourly cumulative emissions for each area. This data is collected through distributed data acquisition terminals and stored according to production unit areas to ensure regional correspondence and temporal continuity.

[0048] Resource efficiency assessment indicators and environmental impact assessment indicators are set based on a multi-source environmental data set. Resource efficiency assessment indicators include the proportion of recycled resources and the clean energy utilization rate; environmental impact assessment indicators include carbon emissions per unit product and waste recycling rate. Initial optimization potential values ​​are generated through a dual-threshold triggering mechanism and a dynamic decay model. The dual-threshold triggering mechanism sets upper and lower thresholds for each assessment indicator; when an indicator exceeds the upper limit or falls below the lower limit, the optimization potential calculation is triggered. The dynamic decay model considers the duration of the indicator's deviation from the threshold; the longer the deviation time, the greater the increase in the initial optimization potential value. For example, if the clean energy utilization rate in the cathode material preparation area is below the lower threshold for three consecutive days, the dynamic decay model will calculate the initial optimization potential value for that area based on the number of consecutive days.

[0049] The initial optimization potential values ​​of each production unit area are dynamically corrected based on the spatial correlation characteristics of adjacent production unit areas, and the corrected optimization potential values ​​are output. Spatial correlation characteristics include material transport relationships, energy sharing, and pollutant diffusion paths. For example, the negative electrode material preparation area and the cell assembly area are connected by a conveyor belt. The waste treatment efficiency of the negative electrode material preparation area affects the environmental quality of the cell assembly area. Therefore, during correction, the initial optimization potential values ​​of the negative electrode material preparation area and the cell assembly area need to be correlated. If there is a significant mutual influence between their waste treatment indicators, their respective initial optimization potential values ​​are adjusted accordingly. For adjacent areas with shared energy pipelines, such as the packaging and testing area and the waste treatment area sharing a solar power supply line, when the optimization potential of clean energy utilization in one area changes, the optimization potential value of the other area also needs to be corrected accordingly to reflect the integrity of the energy system.

[0050] The revised optimization potential values ​​are sorted in descending order to generate a priority ranking table. In the ranking table, production unit areas with higher optimization potential values ​​are listed first. For example, after calculation, the waste treatment area has the highest revised optimization potential value and is ranked first, followed by the cathode material preparation area, etc. This priority ranking table is integrated with the production management system for visualization, including bar charts and heat maps. In the heat map, different production unit areas are displayed with different colors to show their optimization priority; the darker the color, the higher the priority, allowing production managers to intuitively understand the urgency of environmental optimization needs in each area.

[0051] Acquire a dataset of battery manufacturing records for multiple batches corresponding to the target production line. This dataset includes process description text for each batch, time-series data from process sensors, and raw material sustainability indicators. The process description text covers stirring time and temperature settings in the cathode material mixing process, and stacking pressure and alignment accuracy requirements in the cell assembly process. The process sensor time-series data includes real-time rotation speed of the cathode material mixing tank, minute-level readings from temperature sensors, and second-level data from the pressure sensors of the stacking machine. Raw material sustainability indicators include the proportion of recycled cobalt in the cathode material and the biodegradability rate of the separator material.

[0052] Semantic parsing is performed on the textual information describing the process steps, outputting a semantic feature vector. This semantic parsing is achieved using natural language processing (NLP) technology, converting textual information such as "stirring time 30 minutes, temperature 80℃" into vector data containing dimensions such as time and temperature. Each element in the vector corresponds to a standardized value of a process parameter. Fluctuation feature extraction is performed on the time-series data from the process sensors, outputting a dynamic feature vector. Fluctuation features include the average fluctuation amplitude of the sensor data (the average difference between the maximum and minimum values ​​within a certain time period) and fluctuation frequency (the number of times the data crosses the threshold per unit time). These features are arranged chronologically to form the dynamic feature vector. Multimodal feature fusion is performed between the semantic feature vector and the dynamic feature vector to generate a fused production feature map. During fusion, based on the timestamp, the semantic features of the process at the same moment are associated and combined with the dynamic features of the process. For example, the semantic feature (temperature setpoint) at the 5th minute of the stirring process is fused with the sensor dynamic feature (actual temperature fluctuation) at that moment, forming a fused feature that includes the process setting and actual state at that moment. The fused features from multiple moments are arranged chronologically to constitute the fused production feature map.

[0053] A pre-trained anomaly analysis model is invoked to perform anomaly contribution allocation on the fusion manufacturing feature map, generating anomaly correlation distribution maps corresponding to each production batch. The anomaly analysis model, trained on historical production data, can identify anomaly patterns in the fusion manufacturing feature map, such as excessive deviations between the process-set temperature and the actual sensor temperature, or mismatches between material sustainability indicators and process parameters. The anomaly contribution allocation operation calculates the contribution of raw material sustainability indicators and each manufacturing process to the anomaly pattern. For example, when a batch of batteries exhibits carbon emission exceeding standards, the contribution weights of the cathode material recycling ratio, the mixing process, and the sintering process are calculated to form an anomaly correlation distribution map. This map is presented in matrix form, with rows representing raw material sustainability indicators, columns representing manufacturing processes, and matrix element values ​​indicating the anomaly correlation strength between a specific raw material indicator and a specific process.

[0054] Based on the distribution of abnormal correlation strength, a joint root cause analysis is performed on the sustainability indicators of raw materials and the textual descriptions of processes to output the environmental anomaly diagnosis results for each production batch. During the analysis, the focus is on combinations with high anomaly correlation strength. For example, if the correlation strength between the cathode material recycling ratio and the sintering process is the highest, the analysis examines whether the recycling ratio of the cathode material in this batch meets the process requirements. Simultaneously, the temperature curve and holding time of the sintering process are checked for anomalies to pinpoint the source of the problem in the manufacturing process (e.g., incorrect parameter settings in the sintering process) and the type of material defect (e.g., insufficient cathode material recycling ratio). Based on the environmental anomaly diagnosis results, a manufacturing parameter optimization strategy is generated, such as adjusting the temperature setpoint of the sintering process and increasing the proportion of recycled components in the cathode material. This optimization strategy is then transmitted to the production control system, triggering parameter updates for the corresponding equipment and enabling targeted adjustments to the production process.

[0055] Example 5: Obtain the current battery manufacturing parameter configuration set and establish a battery quality and environmental indicator database associated with it. The current battery manufacturing parameter configuration set covers parameters such as material ratio, stirring speed, calcination temperature, and holding time in the electrode preparation stage; stacking pressure, alignment accuracy, and encapsulation strength in the cell assembly stage; and injection volume, injection speed, and vacuum degree in the electrolyte injection stage. The battery quality and environmental indicator database stores quality indicators such as capacity, cycle life, internal resistance, and self-discharge rate of batteries produced under this parameter configuration, as well as environmental indicators such as unit product energy consumption, recycled material utilization rate, carbon emission intensity, and wastewater treatment compliance rate during the production process. Each record in the database contains a corresponding parameter configuration ID, quality indicator value, environmental indicator value, and production batch information. The parameter configuration ID enables the association query between parameters and indicators.

[0056] Based on a database of battery quality and environmental indicators, a set of targeted optimization objectives is configured. This set includes optimization for maximizing capacity, optimizing material renewability, optimizing energy efficiency, and optimizing pollutant emission reduction. Capacity maximization optimization targets the actual discharge capacity of the battery, aiming to increase its capacity output under standard testing conditions. Optimization for material renewability focuses on increasing the proportion of renewable components in raw materials such as cathode materials, anode materials, and separators used in production. Energy efficiency optimization aims to reduce the electrical and thermal energy consumption per unit of battery during production. Pollutant emission reduction optimization aims to reduce the total amount of pollutants such as carbon dioxide, wastewater, and exhaust gases emitted during production. The configuration of this targeted optimization objective set needs to comprehensively consider factors such as the battery's application scenario, market demand, and environmental regulations. For example, high-capacity batteries used in new energy vehicles may require simultaneous emphasis on maximizing capacity and optimizing pollutant emission reduction.

[0057] The optimization priority indicators within the set of targeted optimization objectives are determined. These indicators include primary optimization objectives, secondary optimization objectives, and stability objectives. Primary optimization objectives are those that need to be prioritized within the current production cycle; for example, when market feedback indicates insufficient battery capacity, maximizing capacity becomes the primary objective. Secondary optimization objectives are those that need to be pursued concurrently with the primary objectives; for example, while pursuing maximum capacity, optimizing material recyclability becomes a secondary objective. Stability objectives include the range of battery capacity fluctuations and the standard deviation of production process parameters, ensuring the consistency of battery quality and the stability of the production process during optimization, avoiding excessive batch-to-batch quality differences or drastic fluctuations in equipment operating parameters. The determination of priority indicators requires assessing the interrelationships between the objectives; for example, analyzing whether improving material recyclability will significantly impact battery capacity, and adjusting the priority order based on the degree of impact.

[0058] After selecting the key optimization parameter set using the primary and secondary optimization objectives, a targeted optimization objective impact assessment of the key optimization parameter set is performed. The key optimization parameter set is a combination of parameters that directly impacts the primary and secondary optimization objectives. For example, in a scenario where maximizing capacity is the primary optimization objective and material recyclability is the secondary optimization objective, the key optimization parameter set might include parameters such as the proportion of active material in the cathode material, the doping ratio of recycled material, calcination temperature, and stirring time. The targeted optimization objective impact assessment is achieved by analyzing the degree of influence of changes in each parameter in the key optimization parameter set on the primary and secondary optimization objectives. For example, it assesses the change in battery capacity and the adjustment space for the doping ratio of recycled material when the proportion of active material in the cathode material increases by 1%; it also analyzes the direction and magnitude of the impact on battery capacity and energy consumption when the calcination temperature increases by 5°C. The results of the impact assessment are presented in the form of a parameter-objective impact matrix, where each element represents the numerical impact of a unit change in a parameter on a specific objective, providing a basis for subsequent parameter optimization calculations.

[0059] After establishing parameter optimization constraints based on the impact assessment results of the targeted optimization objectives, parameter-oriented optimization calculations are performed. The results of these calculations are then used to update the environmentally friendly manufacturing parameters for high-capacity batteries. Parameter optimization constraints include the value ranges of each key parameter (such as the upper and lower limits of the calcination temperature) and the correlation constraints between parameters (such as the sum of the active material ratio and the binder ratio in the cathode material being 100%). Parameter-oriented optimization calculations adjust the parameter values ​​in the key optimization parameter set to achieve the preset optimization effects for both the primary and secondary optimization objectives, while ensuring that the stability objective remains within acceptable limits, all while meeting the constraints. For example, in capacity maximization optimization, by gradually adjusting the active material ratio and calcination temperature of the cathode material, the corresponding battery capacity and material recyclability indicators are calculated, and parameter combinations that can both improve capacity and meet material recyclability requirements are selected. The optimized parameter combinations are then transmitted to the production control system to replace the original parameter configuration, updating the environmentally friendly manufacturing parameters for high-capacity batteries. The new parameter configuration will be applied to the next batch of batteries produced.

[0060] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0061] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An environmentally friendly green manufacturing method for a high-capacity battery, characterized in that, include: Data sets of sustainable material properties, green process monitoring data, and environmental footprint indicators related to the battery production process are collected, and initial feature time series are extracted from these data sets. The manufacturing process of the high-capacity battery includes multiple core production stages. Extract a fusion feature set of material-process-environment interaction from the initial feature time series; based on the fusion feature set and the location data acquisition accuracy of different battery manufacturing zones, construct multi-scale model structures with different granularities respectively; and output multi-resolution feature mapping based on the multi-scale model structure generation results. A trend analysis operation is performed on the time evolution data of the multi-resolution feature mapping to generate a battery performance degradation spectrum. Based on the battery performance degradation spectrum, simulation calculation technology is used to determine the stress response distribution of the multiple core production stages and obtain a battery manufacturing state response dataset. Based on the battery manufacturing status response dataset, the number of performance degradation cycles in the core production stage is identified; and by integrating a preset battery material degradation model, the remaining capacity value of the core production stage is calculated. Then, by integrating the remaining capacity values, the overall lifespan of the high-capacity battery in environmentally friendly green manufacturing is evaluated.

2. The environmentally friendly green manufacturing method for a high-capacity battery as described in claim 1, characterized in that, The granularity of the multi-scale model structure is set according to the model unit size parameter, the total number of model elements parameter, and the model density parameter; Based on the fused feature set and combined with the location data acquisition accuracy of different battery manufacturing zones, multi-scale model structures with different granularities are constructed respectively. The multi-scale model structure generates multi-resolution feature maps, including: The battery manufacturing area is divided into several independent sub-manufacturing blocks; Create multiple parallel computing task units, each of which is associated with an independent sub-production block; Based on the location data acquisition accuracy of each independent sub-production block, the model unit size parameters, total number of model elements parameters, and model density parameters of each independent sub-production block are determined; The multiple parallel computing task units are coordinated to generate target multi-scale models for each independent sub-production block based on the model unit size parameters, total number of model elements parameters, and model density parameters of each independent sub-production block. The multi-scale model of the target is integrated with the fused feature set and each independent sub-production block to output the multi-resolution feature map.

3. The environmentally friendly green manufacturing method for a high-capacity battery as described in claim 2, characterized in that, The coordination of the multiple parallel computing task units to generate target multi-scale models for each independent sub-production block based on the model unit size parameters, total number of model elements parameters, and model density parameters of each independent sub-production block includes: The multiple parallel computing task units are controlled to generate preliminary geometric information of the model surface based on the model unit size parameters, total number of model elements parameters, and model density parameters of each independent sub-production block; The multiple parallel computing task units are controlled to construct three-dimensional solid models of each independent sub-production block and obtain model volume topology data based on the geometric information of the model surface; Based on the model topology data, extract the vertex set and volume element set of each 3D entity model; and perform a unique encoding operation on the vertex set and volume element set of each 3D entity model to obtain a unique identification identifier for each vertex set and a unique identification identifier for each volume element set. Control the data synchronization transmission between the multiple parallel computing task units, so that the volume unit sets and exclusive identification marks between adjacent independent sub-production blocks are mutually projected and mapped to generate a projection mapping dataset. Based on the projection mapping dataset, the 3D solid model of each independent sub-production block is adaptively adjusted to obtain the target multi-scale model of each independent sub-production block.

4. The environmentally friendly green manufacturing method for a high-capacity battery as described in claim 1, characterized in that, The extraction of initial feature time series from the sustainable material property dataset, green process monitoring flow data, and environmental footprint indicator information includes: Standardization transformation is performed on sustainable material property datasets, green process monitoring stream data, and environmental footprint index information to output a normalized dataset; A preset feature extraction mechanism is used to extract material chemical features, process dynamic features, and environmental interaction features from a normalized dataset, and these features are then combined to form the initial feature time series.

5. The environmentally friendly green manufacturing method for a high-capacity battery as described in claim 1, characterized in that, The extraction of the fused feature set of material-process-environment interaction from the initial feature time series includes: Extract a primary feature set that couples material properties, process parameters, and environmental factors from the initial feature time series; When the feature dimension of the primary feature group exceeds the preset dimension threshold, a dimensionality reduction operation is performed on the primary feature group to obtain a dimensionality-reduced feature dataset. Correlation analysis is performed on the time series of the dimensionality-reduced feature dataset to enhance the dynamic correlation between material property features, process parameter features, and environmental factor features, thereby generating the fused feature set.

6. The environmentally friendly green manufacturing method for a high-capacity battery as described in claim 1, characterized in that, The high-capacity battery manufacturing process is connected to at least one green energy supply device; The method, based on battery performance degradation spectra, uses simulation calculation techniques to determine the stress response distribution of the multiple core production stages, obtaining a battery manufacturing state response dataset, including: Construct a production system simulation model for the high-capacity battery manufacturing process; construct separate device simulation models for each green energy supply unit, and integrate all device simulation models into the production system simulation model to form a comprehensive simulation model; External excitation input is constructed based on the battery performance degradation spectrum; simulation calculation and analysis are performed based on the comprehensive simulation model and the external excitation input, combined with the preset operating constraints, to obtain the stress response distribution of each core production stage; Based on the stress response distribution of each core production stage, analyze the time variation curves and spatial distribution patterns of stress in each core production stage. By merging the time variation curves and spatial distribution patterns of stress in each core production stage, a dynamic stress evolution map of the high-capacity battery manufacturing process is generated, and a battery manufacturing state response dataset is derived based on the dynamic stress evolution map.

7. The environmentally friendly green manufacturing method for a high-capacity battery as described in claim 1, characterized in that, Also includes: Responding to optimization instructions for a specific production stage, the optimization instructions include parameters for calculating accuracy requirements; Based on the optimization instructions, feature data sequences corresponding to specific production stages are extracted from the battery manufacturing status response dataset. Based on the required accuracy parameters, the performance optimization analysis results for a specific production stage are calculated using feature data sequence analysis.

8. The environmentally friendly green manufacturing method for a high-capacity battery as described in claim 1, characterized in that, Also includes: The battery manufacturing facility is divided into multiple production unit areas. Records of renewable resource consumption, clean energy use data and carbon emission monitoring streams of each production unit area are obtained to form a multi-source environmental data set. Resource efficiency assessment indicators and environmental impact assessment indicators are set based on multi-source environmental protection data sets, and initial optimization potential values ​​are generated through a dual threshold triggering mechanism and a dynamic decay model. Based on the spatial correlation characteristics of adjacent production unit areas, the initial optimization potential value of each production unit area is dynamically corrected, and the corrected optimization potential value is output. The revised optimization potential values ​​are sorted in descending order to generate a priority sorting form, which is then integrated with the production management system for visual display.

9. The environmentally friendly green manufacturing method for a high-capacity battery as described in claim 1, characterized in that, Also includes: Obtain a set of battery manufacturing record data for multiple batches corresponding to the target production line. The set of manufacturing record data includes process description text information, process sensor time series data, and raw material sustainability indicators for each production batch. Semantic parsing is performed on the process description text information to output the process semantic feature vector. Fluctuation feature extraction is performed on the process sensor time series data to output the process dynamic feature vector. The process semantic feature vector and the process dynamic feature vector are fused together to generate a fused production feature map. The pre-trained anomaly analysis model is invoked to perform anomaly contribution allocation on the fusion production feature map, generating anomaly correlation distribution map corresponding to the production batch. The anomaly correlation distribution map includes the distribution of anomaly correlation strength between raw material sustainability indicators and each production process. Based on the distribution of abnormal correlation strength, the joint root cause tracing calculation of the sustainability indicators of raw materials and the textual information of process description is performed to output the environmental anomaly diagnosis results of the production batch. The environmental anomaly diagnosis results are used to locate the production process and material defect type of the source of the problem. Based on the environmental anomaly diagnosis results, a production parameter optimization strategy is generated and transmitted to the production control system to trigger parameter update operations.

10. The environmentally friendly green manufacturing method of a high-capacity battery as described in claim 1, characterized in that, Also includes: Obtain the current battery manufacturing parameter configuration set and establish a database of battery quality and environmental indicators associated with the current battery manufacturing parameter configuration set; A set of targeted optimization objectives is configured based on a database of battery quality and environmental protection indicators. The set of targeted optimization objectives includes capacity maximization optimization, material renewability optimization, energy efficiency optimization, and pollutant emission reduction optimization. Determine the optimization priority identifiers in the set of targeted optimization objectives, wherein the optimization priority identifiers include primary optimization objectives, secondary optimization objectives, and stability objectives; After selecting the key optimization parameter set using the primary and secondary optimization objectives, an impact assessment of the targeted optimization objectives of the key optimization parameter set is performed. After establishing parameter optimization constraints based on the impact assessment results of the targeted optimization objectives, parameter-oriented optimization calculations are performed, and the results of the parameter-oriented optimization calculations are used to update the environmentally friendly manufacturing parameters for high-capacity batteries.

Images