Method for automatically generating battery pack design based on big data intelligent multi-dimensional constraint and generator

Abstract

The invention discloses a design method for automatically generating a battery pack based on big data intelligent multi-dimensional constraint and a generator, and relates to the technical field of battery pack design. According to the design method for automatically generating the battery pack based on the big data intelligent multi-dimensional constraint, a user demand parameter set of the to-be-designed battery pack and a material library stored in a database are input into a pre-trained constraint screening model for comprehensive analysis, and the user demand parameter set of the to-be-designed battery pack and the material library stored in the database are selected within the limited range of material combination; generating one or more initial design schemes meeting user requirements; and based on the electrical data and the manufacturing data of each initial design scheme, analyzing the electrical performance index and the battery pack manufacturing difficulty factor of the corresponding initial design scheme, and generating a comprehensive evaluation optimization index so as to screen the target battery pack design scheme. And the technical requirements of users on the battery pack are met, the feasibility of the manufacturing process is met, and the deliverability of the battery pack product is also realized.

Description

Technical Field

[0001] This invention relates to the field of battery pack design technology, specifically to a method and generator for automatically generating battery pack designs based on big data intelligent multidimensional constraints. Background Technology

[0002] With the continuous development of modern battery technology, especially its widespread application in new energy, smart devices and electric vehicles, battery pack design and optimization have gradually become one of the key technologies. Battery pack design involves many aspects, including but not limited to the selection of individual battery cells, the electrical performance of the battery pack, safety and cost control. Due to the numerous variables and constraints involved in battery pack design, traditional design methods often rely on manual experience and gradual adjustments, which are inefficient and difficult to meet the increasingly complex and diversified market demands.

[0003] To address these issues, technologies such as big data, artificial intelligence, and the Internet of Things have been gradually introduced into battery pack design in recent years. Through big data analysis, combined with real-time monitoring data and historical design data, battery pack design can become more intelligent and efficient. Meanwhile, multi-dimensional constraint optimization methods based on artificial intelligence can automatically generate the optimal design scheme while considering multiple constraints, thereby improving the accuracy and performance of battery pack design.

[0004] Existing technology, such as the patent application with publication number CN114843578B, discloses a visual battery pack design method, including the steps of: receiving the battery pack casing structure dimensions and customized parameters sent by the front end, and generating casing feature interface codes. This invention specifically encodes casing features, cell features, and protection board components, thereby avoiding the cumbersome and cumbersome traditional method of recording component data in a database, and improving the speed of design operations. By setting preset rules, the correlation between various components in the design is standardized, reducing reliance on human experience. Standardized verification testing can reduce the time spent on manual R&D and testing, enabling battery design engineers to quickly design a battery pack that balances safety and performance.

[0005] Based on the above findings, the limitations of existing technologies include at least the following issues: Existing technologies rely on a single-dimensional standard for design evaluation. While this allows for rapid preliminary solutions, it often overlooks multi-dimensional constraints in battery pack design, such as manufacturing difficulty, cost, performance, size, and the safety and long-term stability of the battery pack. Battery pack manufacturing is not simply about assembling cells; it involves complex factors such as the matching of cells with the battery management system, the temperature control design of the battery pack, and the assembly processes of various components. If these factors are not properly evaluated, the design may meet electrical performance requirements but face high costs or production difficulties in actual manufacturing. For example, some high-performance battery pack designs may require extremely complex assembly processes, and existing technologies may struggle to assess these manufacturing difficulties in advance, making large-scale production difficult and ultimately failing to meet user needs. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and generator for automatically generating battery pack designs based on big data intelligent multidimensional constraints. This solves the problem that existing technologies neglect multidimensional constraints, making it difficult to balance performance, cost, and manufacturing feasibility in battery pack design.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for automatically generating battery pack designs based on big data intelligent multi-dimensional constraints, comprising the following steps: obtaining a set of user requirement parameters for the battery pack to be designed, and inputting them into a pre-trained constraint screening model in conjunction with a material library stored in a database for comprehensive analysis, generating several initial design schemes; obtaining electrical data and manufacturing data for each initial design scheme, and obtaining constraint evaluation sets for each initial scheme, including electrical performance indices and battery pack manufacturing difficulty factors; analyzing the comprehensive evaluation optimization index of each initial design scheme based on its constraint evaluation set; and performing screening processing based on the comprehensive evaluation optimization index of each initial design scheme to obtain the target battery pack design scheme.

[0008] Furthermore, the user requirement parameter set includes target voltage value, target capacity value, target size value, target current value, target operating temperature range, and target cost value, and the constraint screening model includes an initial screening sub-network and an intelligent generation sub-network.

[0009] Furthermore, the specific steps for generating several initial design schemes for the battery pack to be designed are as follows: In the initial screening sub-network of the constraint screening model, the user requirement parameter set of the battery pack to be designed and the material library stored in the database are received, and several alternative cell schemes for the battery pack to be designed are analyzed; In the intelligent generation sub-network of the constraint screening model, several initial design schemes for the battery pack to be designed are generated based on each alternative cell scheme for the battery pack to be designed.

[0010] Furthermore, the electrical data includes the total battery voltage, total battery capacity, battery temperature sensitivity factor, battery energy conversion efficiency, battery balance accuracy index, electromagnetic compatibility index, protection circuit response time, and over-temperature protection value. The specific steps for analyzing the electrical performance index of each initial design scheme are as follows: Based on the electrical data of each initial design scheme, analyze the electrical evaluation set of each initial design scheme, including the electrical design accuracy index, performance optimization index, and electrical safety confidence index; based on the electrical evaluation set of each initial design scheme, analyze the electrical performance index of each initial design scheme.

[0011] Furthermore, the specific steps for analyzing the electrical evaluation set of each initial design scheme are as follows: Read the target voltage and target capacity values ​​of the battery pack to be designed, and perform a comprehensive analysis with the total battery voltage and total battery capacity values ​​of each initial design scheme to obtain the electrical design accuracy index of each initial design scheme; based on the battery balance accuracy index, electromagnetic compatibility index, and energy conversion efficiency value of each initial design scheme, analyze the performance optimization index of each initial design scheme; based on the temperature sensitivity factor, protection circuit response time value, and over-temperature protection value of each initial design scheme, analyze the electrical safety confidence index of each initial design scheme.

[0012] Furthermore, the manufacturing data includes thermal expansion factor, contact resistance index, manufacturing process complexity value, production cost redundancy value, material compressive strength value, and thermal diffusivity factor. The specific steps for analyzing the battery pack manufacturing difficulty factor of each initial design scheme are as follows: Based on the manufacturing data of each initial design scheme, analyze the manufacturing feasibility assessment set of each initial design scheme, including the manufacturing process complexity index and structural design reliability index; based on the manufacturing feasibility assessment set of each initial design scheme, analyze the battery pack manufacturing difficulty factor of each initial design scheme.

[0013] Furthermore, the specific steps for analyzing the manufacturing feasibility assessment set of each initial design scheme are as follows: Based on the contact resistance index, manufacturing process complexity value, and production cost redundancy value of each initial design scheme, analyze the manufacturing process complexity index of each initial design scheme; based on the thermal expansion factor, material compressive strength value, and thermal diffusivity factor of each initial design scheme, analyze the structural design reliability index of each initial design scheme.

[0014] Furthermore, the specific calculation of the comprehensive evaluation optimization index for a certain initial design scheme is as follows: Where ZyS is the comprehensive evaluation and optimization index of a certain initial design scheme, DqX is the electrical performance index of a certain initial design scheme, μ1 is the performance coefficient stored in the database, GnY is the battery pack manufacturing difficulty factor of a certain initial design scheme, μ2 is the manufacturing coefficient stored in the database, υ is the proportional coefficient stored in the database, and μ3 is the interaction coefficient stored in the database.

[0015] Furthermore, the specific steps to obtain the target battery pack design scheme are as follows: sort the comprehensive evaluation and optimization index of each initial design scheme in descending order to generate a scheme optimization ranking table; based on the scheme optimization ranking table, analyze the target battery pack design scheme.

[0016] A generator for automatically generating battery pack designs based on big data intelligent multi-dimensional constraints includes: an initial design module, used to acquire the user requirement parameter set of the battery pack to be designed, and input it into a pre-trained constraint screening model for comprehensive analysis in combination with the material library stored in the database, generating several initial design schemes; a constraint evaluation module, used to acquire the electrical and manufacturing data of each initial design scheme, and to obtain the constraint evaluation set of each initial scheme, including electrical performance index and battery pack manufacturing difficulty factor; a comprehensive constraint module, used to analyze the comprehensive evaluation optimization index of each initial design scheme based on the constraint evaluation set; and an optimization screening module, used to perform screening based on the comprehensive evaluation optimization index of each initial design scheme to obtain the target battery pack design scheme.

[0017] The present invention has the following beneficial effects:

[0018] (1) The method for automatically generating battery pack design based on big data intelligent multidimensional constraints, based on user demand parameter set, material library stored in database, and constraint screening model, accurately generates multiple initial design schemes. This ensures that the initial design scheme meets user needs while controlling manufacturing costs to the maximum extent. On this basis, a comprehensive evaluation and optimization index is introduced to screen the initial design scheme, thereby ensuring that the target design scheme not only meets the technical requirements of the battery pack, but can also be smoothly implemented in the manufacturing process, ensuring the feasibility of the production process, thereby effectively reducing manufacturing costs, avoiding redundant expenses, and ensuring smooth production.

[0019] (2) The method of automatically generating battery pack design based on big data intelligent multidimensional constraints enables the battery pack design to be flexibly adjusted according to different user needs through intelligent screening and analysis of multidimensional constraints. Each initial design scheme is based on the user input requirement parameter set and the material library stored in the database. After automatic screening by the constraint screening model, it can quickly adapt to different battery pack design requirements, thereby improving the adaptability of the initial design scheme. It also enables the battery pack to maintain high efficiency under different conditions and ensures that the manufacturing cost is controlled within an acceptable range to meet the actual needs of users.

[0020] (3) The method for automatically generating battery pack design based on big data intelligent multidimensional constraints analyzes the electrical and manufacturing data of each initial design scheme to generate a corresponding comprehensive evaluation and optimization index, thereby accurately identifying the design scheme with the best comprehensive performance in multiple aspects. This can effectively avoid the situation of one-sided optimization of a certain indicator during the design process, and ensure that the battery pack can reach the optimal state in all aspects. For example, the battery pack may focus on voltage or current in terms of electrical performance, but such a design is likely to increase production complexity during the manufacturing process. By using the comprehensive evaluation and optimization index, a balance can be made between multiple dimensions, and finally a design scheme that meets technical requirements and can be successfully mass-produced is selected, thereby reducing the risk in the production process.

[0021] (4) This generator for automatically generating battery pack designs based on big data intelligent multidimensional constraints can quickly generate multiple initial design schemes for battery packs by deeply integrating user requirement parameter sets with material libraries stored in the database. It can also conduct comprehensive analysis and evaluation based on the electrical performance and manufacturing feasibility of each initial design scheme, thereby selecting the optimal design from multiple initial design schemes. Furthermore, it can ensure that the target design scheme meets technical requirements and multidimensional constraints through comprehensive evaluation indicators, thereby improving the automation level of the design process. This can effectively reduce redundant design and cost expenditures, optimize the battery pack production process, improve design efficiency, and ensure the high quality and high stability of the design scheme.

[0022] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0023] Figure 1 This is a flowchart of the method for automatically generating battery pack design based on big data intelligent multidimensional constraints according to the present invention.

[0024] Figure 2 This is a flowchart illustrating the specific steps involved in generating several initial design schemes for a battery pack based on big data intelligent multidimensional constraints in the present invention.

[0025] Figure 3 This is a schematic diagram of the electrical design accuracy index sequence in the method for automatically generating battery pack design based on big data intelligent multidimensional constraints of the present invention.

[0026] Figure 4 This is a schematic diagram of the efficiency optimization index sequence in the method for automatically generating battery pack design based on big data intelligent multidimensional constraints of the present invention.

[0027] Figure 5 This is a schematic diagram of the electrical safety confidence index sequence in the method for automatically generating battery pack design based on big data intelligent multidimensional constraints of the present invention.

[0028] Figure 6 This is a block diagram of the generator for automatically generating battery pack designs based on big data intelligent multidimensional constraints, as per the present invention. Detailed Implementation

[0029] Please see Figure 1 This invention provides a technical solution: a method for automatically generating battery pack designs based on big data intelligent multi-dimensional constraints, comprising the following steps: obtaining a set of user requirement parameters for the battery pack to be designed (by interacting with an e-commerce platform via API, receiving battery pack requirement information from the e-commerce platform online, and extracting the user requirement parameter set based on NLP), and inputting the material library stored in the database (including several types of cells, components, and corresponding attributes such as voltage, current, power, manufacturing cost, etc.) into a pre-trained constraint screening model for comprehensive analysis to generate several initial design schemes; obtaining electrical data and manufacturing data for each initial design scheme, and obtaining constraint evaluation sets for each initial scheme, including electrical performance index and battery pack manufacturing difficulty factor; analyzing the comprehensive evaluation optimization index of each initial design scheme based on the constraint evaluation set of each initial design scheme; and performing screening processing based on the comprehensive evaluation optimization index of each initial design scheme to obtain the target battery pack design scheme (and presenting the target battery pack design scheme as a specification sheet and BOM of the battery pack body (cell assembly), such as listing all components used in the target battery pack design scheme).

[0030] The specific calculation of the comprehensive evaluation and selection index for a certain initial design scheme is as follows: Where ZyS is the comprehensive evaluation and optimization index of a certain initial design scheme, DqX is the electrical performance index of a certain initial design scheme, μ1 is the performance coefficient stored in the database, GnY is the battery pack manufacturing difficulty factor of a certain initial design scheme, μ2 is the manufacturing coefficient stored in the database, υ is the proportional coefficient stored in the database, which is taken as 2.000 in this implementation example, and μ3 is the interaction coefficient stored in the database.

[0031] It needs to be explained that the specific steps for obtaining the performance coefficient μ1 stored in the database are as follows: by obtaining the historical electrical performance index of the historical target design scheme in several historical designs, and extracting the mean and standard deviation of the historical electrical performance index respectively, and performing ratio analysis, the ratio of the standard deviation of the historical electrical performance index to the mean of the historical electrical performance index, and taking the result as the performance coefficient μ1;

[0032] The specific steps for obtaining the manufacturing coefficient μ2 stored in the database are as follows: By obtaining the historical battery pack manufacturing difficulty factor in the historical target design scheme of several historical designs, the mean and standard deviation of the historical battery pack manufacturing difficulty factor are extracted respectively, and a ratio analysis is performed. The ratio of the standard deviation of the historical battery pack manufacturing difficulty factor to the mean of the historical battery pack manufacturing difficulty factor is used as the manufacturing coefficient μ2.

[0033] The specific steps for obtaining the interaction coefficient μ3 stored in the database are as follows: By obtaining the historical electrical performance index and historical battery pack manufacturing difficulty factor from the historical target design schemes of several historical designs, and performing interaction analysis, the interaction values ​​(i.e., historical electrical performance index and historical battery pack manufacturing difficulty factor) in the historical target design schemes of several historical designs are obtained. The historical interaction mean and historical interaction standard deviation are extracted respectively, and a ratio analysis is performed. The historical interaction standard deviation / historical interaction mean is used as the interaction coefficient μ3.

[0034] The specific steps to obtain the target battery pack design scheme are as follows: sort the comprehensive evaluation and optimization index of each initial design scheme in descending order to generate a scheme optimization ranking table; based on the scheme optimization ranking table, analyze the target battery pack design scheme, specifically: select the initial design scheme in the first sequence of the scheme optimization ranking table and mark it as the target battery pack design scheme.

[0035] Specifically, the user requirement parameter set includes target voltage value, target capacity value, target size value, target current (maximum discharge current) value, target operating temperature range, and target cost value. The constraint screening model includes an initial screening sub-network and an intelligent generation sub-network.

[0036] like Figure 2As shown, the specific steps for generating several initial design schemes for the battery pack to be designed are as follows: In the initial screening sub-network of the constraint screening model, the user requirement parameter set of the battery pack to be designed and the material library stored in the database are received, and several alternative cell schemes for the battery pack to be designed are analyzed. Specifically, based on the voltage screening layer, capacity screening layer, and fusion statistics layer, a preliminary screening of multiple cells in the material library stored in the database is performed to generate several alternative cell schemes including series and parallel connections. In the intelligent generation sub-network of the constraint screening model, based on each alternative cell scheme for the battery pack to be designed, several initial design schemes for the battery pack to be designed are generated. Specifically, based on the topology generation layer, component matching layer, and cost verification layer, each alternative cell scheme is screened again, and suitable components are assembled to generate several initial design schemes.

[0037] The initial screening subnetwork includes a voltage screening layer, a capacity screening layer, and a fusion statistics layer. The fusion screening layer includes a cost screening layer, a current screening layer, an operating temperature screening layer, and a size screening layer.

[0038] The voltage screening layer reads the voltage value of each type of battery cell from the material library stored in the database and performs a matching analysis with the target voltage value to be designed. Obtain the number of cells connected in series for each type of battery cell;

[0039] The capacity screening layer reads the capacity value of each type of battery cell from the material library stored in the database and performs a matching analysis with the target capacity value to be designed. Obtain the number of each type of battery cell connected in parallel;

[0040] The cost screening layer reads the series and parallel quantity values ​​of each type of battery cell, multiplies them to obtain the required quantity value of each type of battery cell, and reads the manufacturing cost value of each type of battery cell from the material library stored in the database, multiplies it to obtain the budgeted cost value of each type of battery cell, and determines whether it is lower than the target cost value. If it is lower, the battery cell is retained (marked as the first usable battery cell); otherwise, it is discarded.

[0041] The current screening layer reads the current value of each type of first available cell from the material library stored in the database and checks whether it is higher than or equal to the target current value. If it is higher than or equal to the target current value, the cell is retained (marked as the second available cell); otherwise, it is discarded.

[0042] The working temperature screening layer reads the working temperature range of each second available battery cell stored in the material library in the database and determines whether it is within the target working temperature range. If it is within the target working temperature range, the battery cell is retained (marked as the third available battery cell); otherwise, it is discarded.

[0043] The size screening layer reads the dimensions (length, width, and height) of each third available cell from the material library stored in the database. Based on the series and parallel connection values ​​of each third available cell, it performs size analysis to generate the size value of each third available cell (i.e., length × series connection value, width × parallel connection value, and height × parallel connection value). It then determines whether the size is lower than the target size. If it is, the cell is retained (marked as a fourth available cell); otherwise, it is discarded. Based on the series and parallel connection values ​​required for each fourth available cell, several alternative schemes for the battery pack to be designed are obtained.

[0044] The intelligent generation subnetwork includes a topology generation layer, a component matching layer, and a cost verification layer.

[0045] The topology generation layer reads several alternative schemes for the battery pack to be designed. Based on the number of cells in series for each alternative scheme, it connects the cells in series sequentially to generate multiple series units. Based on the number of cells in parallel for each alternative scheme, it connects the multiple series units in parallel to ensure the accumulation of current and capacity. Assuming the battery pack needs to be designed as 4 series and 5 parallel: each series unit contains 4 cells, and 5 sets of series units are connected in parallel. The cell topology diagram of the battery pack will show the 4 series and 5 parallel connection method and list the electrical connection points of all cells in the battery pack in this alternative scheme, including series connection points, parallel connection points, and output ports, such as positive connection point: the positive port of the battery pack (connected to the positive current output terminal of the battery pack), negative connection point: the negative port of the battery pack (connected to the negative current output terminal of the battery pack), series connection node: the connection point of each series unit, and parallel connection node: the connection point of each parallel unit.

[0046] The component matching layer, based on the generation results of the topology generation layer, matches a battery management system (BMS) for each alternative scheme. When selecting a BMS module, it determines whether the voltage range of the BMS module stored in the database is higher than the total voltage of the alternative scheme (number of series connections × cell voltage value in the alternative scheme) and whether the current range is higher than the maximum discharge current of the battery pack (cell current value in the alternative scheme × number of parallel connections). If it is higher than both the total voltage and the maximum discharge current of the alternative scheme, then it is assembled with the alternative scheme.

[0047] Matching current protection circuit layers (such as fuses and connectors): Fuse selection rules: The fusing current of the fuse should be 1.5 times the target current (using a safety factor to design the fuse fusing current), and the withstand voltage of the fuse should be greater than or equal to the total voltage of the battery pack; Connector selection rules: The withstand voltage of the connector should be higher than the total voltage of the battery pack, and the current carrying capacity of the connector (the maximum current that the connector can safely pass through) should be higher than the maximum discharge current of the battery pack. If both the fuse selection rules and the connector selection rules are met, then it should be assembled with this alternative solution.

[0048] The cost verification layer reads the manufacturing cost of each cell and required components for each alternative scheme, sums them up, and determines whether it is lower than the target cost value. If it is lower, the alternative scheme is retained and marked as the initial design scheme. This includes several components and corresponding BMS and current protection circuits, such as grouping method (series and parallel), appearance dimensions, etc. It also includes the key parameters of the battery pack body (cell assembly) in the corresponding initial design scheme, which are obtained by obtaining the specific values ​​of the components corresponding to the initial design scheme from the material library stored in the database.

[0049] The pre-training steps for the constraint screening model are as follows:

[0050] In the pre-training stage of the constraint screening model, a training dataset covering multi-dimensional battery design features is first constructed. Specifically, this includes collecting a historical battery pack design scheme library (covering typical application scenarios such as electric vehicles, energy storage systems, and consumer electronics), integrating anonymized real user demand records from e-commerce platforms, and importing full parameters from the material library (including basic attributes such as cell voltage / current / capacity / size, component technical specifications, and supply chain cost data).

[0051] For the above data types, feature engineering processing is performed as follows: user requirement parameters are vectorized and encoded to generate normalized feature vectors; a cell compatibility matrix is ​​constructed for the material library to calculate the voltage matching degree, capacity matching degree, and cost deviation of each cell model; the topology is converted into connection relationship data that can be processed by a graph neural network, where cells are nodes and electrical connections are edges; and an association knowledge graph is established for component matching relationships to form a multi-level mapping relationship of BMS model, voltage range, and cell type.

[0052] After completing the training data preparation, the two sub-networks of the constraint screening model were pre-trained in stages. Taking the initial screening sub-network as an example, the initial screening sub-network was trained independently using supervised learning. Its input was the concatenated tensor of the user demand vector and the cell feature matrix, and the output was the confidence score of the candidate cell scheme. During the training process, the network parameters were optimized hierarchically: the voltage screening layer learned the prediction of the number of series connections through the mean square error loss function (calculating the square difference between the predicted value and the true value), the capacity screening layer learned the prediction of the number of parallel connections through the same mechanism, and the fusion statistics layer (including cost / current / temperature / size screening) used the cross-entropy loss function to evaluate the feasibility of the scheme. The training used the AdamW optimizer (learning rate 0.001, weight decay 1e-4), and a hard sample mining strategy was introduced to strengthen the learning of critical cases of cost and size.

[0053] After the sub-networks are pre-trained independently, end-to-end fusion optimization is performed: a differentiable interface is established between the initial screening sub-network and the intelligent generation sub-network, so that the scheme feasibility score gradient can be backpropagated to the screening threshold parameter. Constraint conflicts are dynamically balanced through a multi-objective optimization algorithm to construct a function that maximizes benefits (an optimization objective that combines the inverse of cost and performance weighting). At the same time, hard constraints such as size upper limit and temperature range are applied. Finally, the deployment architecture receives user requirements through API. After requirement parsing and processing, the constraint screening model drives the scheme generation engine to output directly producible 3D design files and BOM lists.

[0054] In this implementation plan, meticulous constraint screening and intelligent generation processes effectively improve the accuracy and automation of battery pack design. First, through rapid analysis of user requirement parameter sets and comprehensive evaluation of the material library stored in the database, multiple design schemes can be generated quickly, ensuring that each initial design scheme meets user requirements and minimizes costs. Second, through steps such as topology generation and component matching, the cell combination and component selection of the battery pack are ensured to conform to the technical requirements of the battery pack, thereby avoiding unreasonable design schemes and preventing battery pack performance degradation or non-compliant production due to mismatched components. This achieves the best balance between cost and performance while effectively avoiding redundant design.

[0055] Specifically, the electrical data includes the total battery voltage (obtained by multiplying the number of series connections selected in the initial design scheme by the corresponding cell voltage), the total battery capacity (obtained by multiplying the number of parallel connections selected in the initial design scheme by the corresponding cell capacity), battery temperature sensitivity factor, battery energy conversion efficiency, battery balance accuracy index, electromagnetic compatibility index, protection circuit response time, and over-temperature protection value. The specific steps for analyzing the electrical performance index of each initial design scheme are as follows: Based on the electrical data of each initial design scheme, analyze the electrical evaluation set of each initial design scheme, including the electrical design accuracy index, efficiency optimization index, and electrical safety confidence index; Based on the electrical evaluation set of each initial design scheme, analyze the electrical performance index of each initial design scheme (used to screen the key parameters of the battery pack body in the initial design scheme to achieve the optimal performance of each key parameter while meeting the requirements).

[0056] Among them, the temperature sensitivity factor is the performance change of the battery under different operating temperatures. It can be obtained by averaging the temperature sensitivity values ​​of each cell in the material library stored in the database.

[0057] The battery balance accuracy index is the accuracy with which the BMS achieves battery pack balance during charging. It is obtained by using the balance accuracy of the BMS selected in the initial design scheme from the material library stored in the database.

[0058] The energy conversion efficiency value is the energy conversion efficiency of the battery during the charging and discharging process of the battery pack in this initial design scheme. The charging and discharging efficiency values ​​of each cell in this initial design scheme are obtained by obtaining the material library stored in the database and performing weighted averaging to obtain the energy conversion efficiency value.

[0059] The electromagnetic compatibility index is the electromagnetic shielding performance of the enclosure, which prevents electromagnetic interference. It is obtained from the shielding effectiveness value (the enclosure's ability to prevent electromagnetic interference) of the enclosure selected in the initial design scheme stored in the material library in the database.

[0060] The protection circuit response time is the time required for the protection circuit to detect an anomaly (such as overcurrent or overtemperature) and take action (such as disconnecting the circuit or shutting off the power). It is obtained from the protection circuit selected in the initial design scheme from the material library stored in the database.

[0061] The over-temperature protection value is the cell temperature when the protection circuit is activated, which is obtained from the protection value of the protection circuit selected in the initial design scheme stored in the material library in the database.

[0062] The specific steps for calculating the electrical performance index of an initial design scheme are as follows: Wherein, DqX is the electrical performance index of a certain initial design scheme, SdJ is the electrical design accuracy index of a certain initial design scheme, α1 is the accuracy coefficient stored in the database, XnH is the efficiency optimization index of a certain initial design scheme, α2 is the efficiency coefficient stored in the database, DaK is the electrical safety confidence index of a certain initial design scheme, and α3 is the confidence coefficient stored in the database. The adjustment coefficient is stored in the database and is set to 3.000 in this implementation example.

[0063] It needs to be explained that the steps for obtaining the accuracy coefficient α1 stored in the database are as follows: read the target voltage value, target capacity value, total battery voltage value, and total battery capacity value of the initial design scheme, and perform ratio analysis (such as total voltage value / target voltage value) to obtain the voltage ratio and capacity ratio of the initial design scheme. Then, perform weighted processing, and the result is the accuracy coefficient α1.

[0064] The steps to obtain the efficiency coefficient α2 stored in the database are as follows: Read the battery balance accuracy index, electromagnetic compatibility index, and energy conversion efficiency value of the initial design scheme and perform standardization processing. Based on the standardization processing results, extract the mean efficiency and standard deviation efficiency of the initial design scheme respectively, and perform ratio processing, that is, standard deviation efficiency / mean efficiency. The result is the efficiency coefficient α2.

[0065] The steps to obtain the confidence coefficient α3 stored in the database are as follows: Read the temperature sensitivity factor, protection circuit response time value, and over-temperature protection value of the initial design scheme, and perform standardization processing. Based on the standardization processing results, extract the confidence mean and confidence standard deviation of the initial design scheme respectively, and perform ratio processing, that is, confidence standard deviation / confidence mean. The result is the efficiency coefficient α3.

[0066] The following is a specific implementation example for calculating the electrical performance index of an initial design scheme, including the following data: electrical design accuracy index, efficiency optimization index, and electrical safety confidence index of five initial design schemes, as shown in Table 1 and... Figure 3-5 As shown:

[0067] Table 1. Example of electrical assessment set sequence data

[0068] Electrical design accuracy index Performance Optimization Index Electrical safety confidence index Initial Design Scheme 1 0.876 0.816 0.716 Initial Design Scheme 2 0.713 0.793 0.764 Initial Design Scheme 3 0.839 0.776 0.812 Initial design scheme 4 0.738 0.836 0.843 Initial design scheme 5 0.691 0.784 0.728

[0069] The precision coefficient α1 stored in the database is approximately 0.764;

[0070] The performance coefficient α2 stored in the database is approximately 0.438;

[0071] The confidence coefficient α3 stored in the database is approximately 0.236;

[0072] Adjustment coefficients stored in the database The value is: 3.000;

[0073] Substituting the data from Table 1 and the coefficients mentioned above into the specific formula for calculating the electrical performance index of a certain initial design scheme, we obtain:

[0074] The electrical performance index of the first initial design scheme is approximately 0.704, calculated as follows: (√(ln(1+0.764×0.876))+0.816^0.438+0.236×exp0.716) / 3.000.

[0075] The electrical performance index of the second initial design scheme is approximately 0.689, calculated as follows: (√(ln(1+0.764×0.713))+0.793^0.438+0.236×exp0.764) / 3.000.

[0076] The electrical performance index of the third initial design scheme is approximately 0.709, calculated as follows: (√(ln(1+0.764×0.839))+0.776^0.438+0.236×exp0.812) / 3.000.

[0077] The electrical performance index of the fourth initial design scheme is approximately 0.714, calculated as follows: (√(ln(1+0.764×0.738))+0.836^0.438+0.236×exp0.843) / 3.000.

[0078] The electrical performance index of the fifth initial design scheme is approximately 0.679, calculated as follows: (√(ln(1+0.764×0.691))+0.836^0.784+0.236×exp0.728) / 3.000.

[0079] The specific steps for analyzing the electrical evaluation set of each initial design scheme are as follows: Read the target voltage and target capacity values ​​of the battery pack to be designed, and perform a comprehensive analysis with the total battery voltage and total battery capacity values ​​of each initial design scheme to obtain the electrical design accuracy index for each initial design scheme. Specifically, for each initial design scheme, the total battery voltage and total battery capacity values ​​are compared with the target voltage and target capacity values ​​of the battery pack to be designed (e.g., total voltage value - target voltage value, and the reciprocal of the result). A weighted average is then performed based on the difference results, and the resulting value is the electrical design accuracy index. The battery balance accuracy index for each initial design scheme is then calculated. The efficiency optimization index of each initial design scheme is analyzed based on the battery balance accuracy index, electromagnetic compatibility index, and energy conversion efficiency value. Specifically, the battery balance accuracy index, electromagnetic compatibility index, and energy conversion efficiency value are standardized, and weighted processing is performed based on the standardized results. The result is the efficiency optimization index. The electrical safety confidence index of each initial design scheme is analyzed based on the temperature sensitivity factor, protection circuit response time value, and over-temperature protection value. Specifically, the temperature sensitivity factor, protection circuit response time value, and over-temperature protection value are standardized, and weighted processing is performed based on the standardized results. The result is the electrical safety confidence index.

[0080] In this implementation plan, by integrating multiple electrical data and evaluation indicators, the design scheme of the battery pack can be accurately analyzed and optimized to ensure that the final target technical solution has good electrical performance. First, through detailed analysis of electrical data, the electrical performance index of each initial design scheme is obtained. Then, through standardization processing and weighted analysis, multiple evaluation results are generated, which helps to quickly screen out the design scheme that meets the technical requirements and has high efficiency. Second, multiple data processing and optimization algorithms are adopted to maximize the safety of the battery pack. Furthermore, through accurate electrical performance indices, the best-performing scheme is quickly screened, thereby avoiding non-compliant or inefficient designs and reducing the risk of unqualified products.

[0081] Specifically, the manufacturing data includes thermal expansion factor, contact resistance index, manufacturing process complexity value, production cost redundancy value, material compressive strength value, and thermal diffusivity factor. The specific steps for analyzing the battery pack manufacturing difficulty factor of each initial design scheme are as follows: Based on the manufacturing data of each initial design scheme, analyze the manufacturing feasibility assessment set of each initial design scheme, including the manufacturing process complexity index and the structural design reliability index; Based on the manufacturing feasibility assessment set of each initial design scheme, analyze the battery pack manufacturing difficulty factor of each initial design scheme, which specifically involves weighting the manufacturing process complexity index and the structural design reliability index of each initial design scheme, and the result is the battery pack manufacturing difficulty factor.

[0082] It should be noted that the weighting coefficients for each index in the weighted processing involved in this implementation example are all obtained through the particle swarm optimization algorithm. Taking the specific steps of obtaining the weighting coefficients for the manufacturing complexity index and structural design reliability index of a certain initial design scheme as an example, in the particle swarm optimization algorithm, each particle represents a possible solution, and each solution consists of the weighting coefficients corresponding to the manufacturing complexity index and structural design reliability index. Initially, the position and velocity of the particles are randomly generated, and these positions correspond to different weight combinations. The fitness of each particle is evaluated through an objective function, which is related to the optimization of the battery pack manufacturing yield and production efficiency. Specifically, the objective function is calculated... The particle swarm optimization (PSO) algorithm calculates the mean squared error between the predicted and actual manufacturing difficulty factors. It measures the quality of each solution by calculating the objective function value; a smaller objective function value indicates a better solution. Based on the fitness function value, the PSO algorithm updates the particle's velocity and position. Particles adjust their velocity according to their historical and global best positions, thus searching the solution space and gradually approaching the optimal solution. After multiple iterations, the algorithm gradually converges until a termination condition is met, such as reaching the maximum number of iterations or the fitness change being less than a preset threshold. After multiple iterations and updates, the particles eventually find a set of optimal solutions, namely the weight coefficients corresponding to the manufacturing complexity index and the structural design reliability index.

[0083] The thermal expansion factor is the degree of structural deformation during the assembly of each component (such as battery cell, casing, etc.) in the initial design scheme. The thermal expansion coefficient of each component is obtained by obtaining the material library stored in the database and averaging it. The result is the thermal expansion factor.

[0084] The contact resistance index is the contact resistance between the battery cells and other components (such as the battery management system, connectors, etc.) inside the battery pack. It is obtained by acquiring the resistivity of each cell and contact component (resistivity of the contact component material) from the material library stored in the database, and obtaining the distance and contact area between each cell and contact component in the initial design scheme (which can be automatically extracted by using CAD tools such as SolidWorks, AutoCAD, CATIA, etc. to create a 3D model of the battery pack). The ratio is then processed as (resistivity × distance) / contact area, and the average value is calculated based on the ratio processing result.

[0085] The production process complexity value is the complexity of the production process in the initial design scheme. The production step value and the difficulty factor of the corresponding production step are obtained by obtaining the production time value of each step in the past several production times and taking the reciprocal. The weighted summation is then performed to obtain the step complexity factor of each component. Finally, the weighted average is performed, and the result is the production process complexity value.

[0086] Production cost redundancy is the degree of difference between the cost incurred in the initial design and the target cost, representing the degree of fault tolerance in the manufacturing process. It is calculated by summing the manufacturing costs of each component in the material library stored in the database, obtaining the total production cost, and then comparing it to the target cost: (target cost - total production cost) / target cost. This ratio yields the production cost redundancy.

[0087] The compressive strength value of a material is the compressive strength of the material during the assembly of each component (such as the battery cell, casing, etc.) in the initial design scheme. The compressive strength value of each component is obtained from the material library stored in the database and then averaged. The result is the compressive strength value of the material.

[0088] The thermal diffusivity factor is the degree of thermal diffusion during the assembly of each component (such as battery cell, casing, etc.) in the initial design scheme. The thermal diffusivity coefficient of each component is obtained from the material library stored in the database and then averaged. The result is the thermal diffusivity factor.

[0089] The specific steps for analyzing the manufacturing feasibility assessment set of each initial design scheme are as follows: Based on the contact resistance index, manufacturing process complexity value, and production cost redundancy value of each initial design scheme, the manufacturing process complexity index of each initial design scheme is analyzed. Specifically, the contact resistance index, manufacturing process complexity value, and production cost redundancy value of each initial design scheme are standardized, and the results of the standardization are weighted to obtain the manufacturing process complexity index. Based on the thermal expansion factor, material compressive strength value, and thermal diffusivity factor of each initial design scheme, the structural design reliability index of each initial design scheme is analyzed. Specifically, the thermal expansion factor, material compressive strength value, and thermal diffusivity factor of each initial design scheme are standardized, and the results of the standardization are weighted to obtain the structural design reliability index.

[0090] This implementation scheme comprehensively considers multiple manufacturing-related parameters, enabling precise assessment of the manufacturing difficulty of each initial design scheme. It integrates considerations such as the production difficulty and cost control of battery packs, ensuring that the initial design schemes not only meet performance requirements but also possess low complexity and high feasibility during the production process. Secondly, the introduction of particle swarm optimization algorithm ensures an optimal balance between the complexity of the production process and the reliability of the design, effectively avoiding overly complex or unrealistic design schemes, thereby improving production efficiency and reducing production costs. Finally, through standardization and weighted processing methods, imbalances and deviations in various data can be eliminated, ensuring the quality of the final design scheme and smooth production, thus effectively improving the manufacturability of the battery pack design scheme and guaranteeing its high efficiency in actual production.

[0091] Please see Figure 6 This invention provides a technical solution: a generator for automatically generating battery pack designs based on big data intelligent multi-dimensional constraints, comprising: an initial design module, used to acquire a set of user requirement parameters for the battery pack to be designed, and input them into a pre-trained constraint screening model in conjunction with a material library stored in a database for comprehensive analysis, generating several initial design schemes; a constraint evaluation module, used to acquire electrical data and manufacturing data for each initial design scheme, and to obtain a constraint evaluation set for each initial scheme, including an electrical performance index and a battery pack manufacturing difficulty factor; a comprehensive constraint module, used to analyze the comprehensive evaluation optimization index of each initial design scheme based on the constraint evaluation set; and an optimization screening module, used to perform screening processing based on the comprehensive evaluation optimization index of each initial design scheme to obtain a target battery pack design scheme.

[0092] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0093] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for automatically generating battery pack designs based on big data intelligent multidimensional constraints, characterized in that, Includes the following steps: Obtain the user requirement parameter set of the battery pack to be designed, and input it into the pre-trained constraint screening model in combination with the material library stored in the database for comprehensive analysis to generate several initial design schemes. Obtain electrical and manufacturing data for each initial design scheme, and establish a constraint evaluation set for each initial scheme, including electrical performance index and battery pack manufacturing difficulty factor. Based on the constraint evaluation set of each initial design scheme, analyze the comprehensive evaluation optimization index of the corresponding initial design scheme; The target battery pack design scheme is obtained by screening based on the comprehensive evaluation and optimization index of each initial design scheme.

2. The method for automatically generating battery pack design based on big data intelligent multidimensional constraints according to claim 1, characterized in that, The user requirement parameter set includes target voltage value, target capacity value, target size value, target current value, target operating temperature range, and target cost value. The constraint screening model includes an initial screening sub-network and an intelligent generation sub-network.

3. The method for automatically generating battery pack design based on big data intelligent multidimensional constraints according to claim 2, characterized in that, The specific steps for generating several initial design schemes for the battery pack to be designed are as follows: In the initial screening sub-network of the constraint screening model, the user requirement parameter set of the battery pack to be designed and the material library stored in the database are received, and several alternative cell schemes of the battery pack to be designed are analyzed. In the intelligent generation subnetwork of the constraint screening model, several initial design schemes for the battery pack to be designed are generated based on each alternative cell scheme of the battery pack to be designed.

4. The method for automatically generating battery pack design based on big data intelligent multidimensional constraints according to claim 1, characterized in that, The electrical data includes total battery voltage, total battery capacity, battery temperature sensitivity factor, battery energy conversion efficiency, battery balance accuracy index, electromagnetic compatibility index, protection circuit response time, and over-temperature protection value. The specific steps for analyzing the electrical performance index of each initial design scheme are as follows: Based on the electrical data of each initial design scheme, analyze the electrical evaluation set of each initial design scheme, including the electrical design accuracy index, performance optimization index, and electrical safety confidence index; Based on the electrical evaluation set of each initial design scheme, the electrical performance index of each initial design scheme is analyzed.

5. The method for automatically generating battery pack design based on big data intelligent multidimensional constraints according to claim 4, characterized in that, The specific steps for analyzing the electrical evaluation set of each initial design scheme are as follows: Read the target voltage and target capacity values ​​of the battery pack to be designed, and perform a comprehensive analysis with the total battery voltage and total battery capacity values ​​of each initial design scheme to obtain the electrical design accuracy index of each initial design scheme; Based on the battery balance accuracy index, electromagnetic compatibility index, and energy conversion efficiency value of each initial design scheme, the performance optimization index of each initial design scheme is analyzed. Based on the temperature sensitivity factor, protection circuit response time value, and over-temperature protection value of each initial design scheme, the electrical safety confidence index of each initial design scheme is analyzed.

6. The method for automatically generating battery pack design based on big data intelligent multidimensional constraints according to claim 1, characterized in that, The manufacturing data includes thermal expansion factor, contact resistance index, manufacturing process complexity value, production cost redundancy value, material compressive strength value, and thermal diffusivity factor. The specific steps for analyzing the battery pack manufacturing difficulty factor of each initial design scheme are as follows: Based on the manufacturing data of each initial design scheme, analyze the manufacturing feasibility assessment set of each initial design scheme, including the process manufacturing complexity index and the structural design reliability index; Based on the manufacturing feasibility assessment set of each initial design scheme, the battery pack manufacturing difficulty factor of each initial design scheme is analyzed.

7. The method for automatically generating battery pack design based on big data intelligent multidimensional constraints according to claim 6, characterized in that, The specific steps for analyzing the manufacturing feasibility assessment set of each initial design scheme are as follows: Based on the contact resistance index, manufacturing process complexity value, and production cost redundancy value of each initial design scheme, the manufacturing process complexity index of each initial design scheme is analyzed. Based on the thermal expansion factor, material compressive strength, and thermal diffusivity of each initial design scheme, the structural design reliability index of each initial design scheme is analyzed.

8. The method for automatically generating battery pack design based on big data intelligent multidimensional constraints according to claim 1, characterized in that, The specific calculation of the comprehensive evaluation and selection index for a certain initial design scheme is as follows: Among them, ZyS, DqX, and GnY are the comprehensive evaluation and optimization index, electrical performance index, and battery pack manufacturing difficulty factor of a certain initial design scheme, respectively, and μ1, μ2, υ, and μ3 are the performance coefficient, manufacturing coefficient, proportional coefficient, and interaction coefficient stored in the database, respectively.

9. The method for automatically generating battery pack design based on big data intelligent multidimensional constraints according to claim 1, characterized in that, The specific steps to obtain the target battery pack design scheme are as follows: The comprehensive evaluation and selection index of each initial design scheme is sorted in descending order to generate a scheme selection ranking table; Based on the optimal ranking table, the design scheme of the target battery pack is analyzed.

10. A generator for automatically generating battery pack designs based on big data intelligent multidimensional constraints, using the method for automatically generating battery pack designs based on big data intelligent multidimensional constraints as described in any one of claims 1-9, characterized in that, include: The initial design module is used to obtain the user requirement parameter set of the battery pack to be designed, and combine it with the material library stored in the database to input into the pre-trained constraint screening model for comprehensive analysis, generating several initial design schemes. The constraint evaluation module is used to acquire electrical and manufacturing data for each initial design scheme and to generate constraint evaluation sets for each initial scheme, including electrical performance index and battery pack manufacturing difficulty factor. The comprehensive constraint module is used to analyze the comprehensive evaluation optimization index of the corresponding initial design scheme based on the constraint evaluation set of each initial design scheme; The optimization and screening module is used to screen based on the comprehensive evaluation and optimization index of each initial design scheme to obtain the target battery pack design scheme.

Images