Digital twin method and system for flexible brushless motor production line

Abstract

The present invention provides a flexible brushless motor production line digital twin method and system thereof, which relate to the field of intelligent manufacturing technology, including collecting a multi-source data set from the flexible brushless motor production line, and extracting timing features from the multi-source data set; using a recursive neural network to analyze the extracted timing features, and outputting a rule adjustment strategy; based on the rule adjustment strategy, updating a preset data mapping rule in real time to obtain an updated data mapping rule; according to the updated data mapping rule, filling the extracted timing features into a plurality of pre-built simulation modules; based on the time sequence determined by the timing features of each simulation module, running corresponding and pre-configured simulation modules on each distributed computing node of a pre-built distributed computing framework, so as to simulate the production status of the flexible brushless motor production line in real time and realize the digital twin of the flexible brushless motor production line.

Description

Technical Field

[0001] The present invention relates to the field of intelligent manufacturing technology, and in particular to a digital twin method and system for a flexible brushless motor production line. Background Art

[0002] With the intelligent transformation of the manufacturing industry, flexible production line technology has become an important means to improve production efficiency, reduce costs and achieve customized production. In the field of motor manufacturing, flexible brushless motor production lines have gradually become the mainstream of the market due to their advantages such as high efficiency, low noise and long life. However, how to efficiently manage and optimize such complex production lines has become a problem that needs to be solved urgently.

[0003] Traditionally, the management and optimization of production lines mainly rely on manual experience, which is not only time-consuming and labor-intensive, but also difficult to ensure accuracy and real-time performance. With the rapid development of the Internet of Things, big data, and artificial intelligence technologies, digital twin technology has emerged, providing a new solution for the intelligent management and optimization of production lines. Digital twin technology establishes a virtual model of the production line to reflect the actual operating status of the production line in real time, thereby achieving comprehensive monitoring and precise optimization of the production process.

[0004] Therefore, it is necessary to provide a digital twin method and system for a flexible brushless motor production line to solve the problems existing in traditional production lines. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a digital twin method and system for a flexible brushless motor production line. By collecting multi-source data sets on the flexible brushless motor production line, timing features are extracted, and the timing features are analyzed using a recursive neural network to output a rule adjustment strategy; based on the rule adjustment strategy, the preset data mapping rules are updated in real time, and the extracted timing features are correspondingly filled into multiple pre-built simulation modules; finally, the simulation module is run on each distributed computing node of the distributed computing framework to simulate the production status of the flexible brushless motor production line in real time to realize the digital twin.

[0006] The present invention provides a digital twin method for a flexible brushless motor production line, the method comprising the following steps:

[0007] Collecting a multi-source data set from a flexible brushless motor production line, and extracting time series features from the multi-source data set;

[0008] Using a recursive neural network to analyze the extracted time series features and output a rule adjustment strategy, wherein the recursive neural network is used to identify change characteristics of the time series features;

[0009] Based on the rule adjustment strategy, the preset data mapping rule is updated in real time to obtain an updated data mapping rule, wherein the preset data mapping rule is a predefined feature classification rule;

[0010] According to the updated data mapping rules, the extracted timing features are correspondingly filled into a plurality of pre-built simulation modules, wherein the simulation modules are a plurality of independent but interrelated functional units pre-defined according to the production process of the flexible brushless motor production line;

[0011] Based on the time sequence determined by the timing characteristics of each simulation module, the corresponding and pre-configured simulation modules are run on each distributed computing node of the pre-built distributed computing framework to simulate the production status of the flexible brushless motor production line in real time and realize the digital twin of the flexible brushless motor production line.

[0012] Preferably, the collecting of multi-source data sets from a flexible brushless motor production line and extracting timing features from the multi-source data sets include:

[0013] Capturing the operation data, equipment status data, quality control data and environmental parameter data on the flexible brushless motor production line in real time through sensor networks and Internet of Things technologies to form the multi-source data set;

[0014] Preprocessing the multi-source data set, wherein the preprocessing includes removing noise data and irrelevant features;

[0015] A feature extraction algorithm is used to extract time series features from the preprocessed multi-source data set, wherein the time series features include peak features, slope features, statistics features, spectrum density features, phase features and transient behavior features.

[0016] Preferably, the recursive neural network is used to analyze the extracted time series features and output a rule adjustment strategy, wherein the recursive neural network is used to identify the change characteristics of the time series features, including:

[0017] The extracted time series features are used as input data into a pre-built and trained recurrent neural network model;

[0018] The recursive neural network model captures the time dependency and long-term dependency in the time series features through its internal recursive structure to identify the change characteristics of the time series features, wherein the change characteristics include trend changes, periodic fluctuations and abnormal events;

[0019] Based on the identified change characteristics of the timing characteristics, the recursive neural network model outputs a rule adjustment strategy, and the rule adjustment strategy is used to guide the real-time update of the preset data mapping rules to optimize the simulation accuracy of the simulation module on the production status of the flexible brushless motor production line.

[0020] Preferably, the recursive neural network model output rule adjustment strategy based on the identified change characteristics of the time series characteristics includes:

[0021] Based on the change characteristics of the identified timing characteristics, determine the degree of impact and potential impact range on the production status of the flexible brushless motor production line;

[0022] According to the degree of impact and potential impact range, the recursive neural network model automatically calculates and generates an initial rule adjustment strategy;

[0023] The generated initial rule adjustment strategy is compared and verified with the preset strategy evaluation criteria, and the final rule adjustment strategy is determined based on the comparison and verification results.

[0024] Preferably, the updating of the preset data mapping rules in real time based on the rule adjustment strategy to obtain the updated data mapping rules includes:

[0025] Receiving a rule adjustment strategy from a recursive neural network model, wherein the rule adjustment strategy adjusts a feature classification threshold, a feature classification category, or a feature classification weight in a preset data mapping rule;

[0026] Parse the preset data mapping rules and identify the feature classification parameters that need to be adjusted;

[0027] According to the received rule adjustment strategy, the identified feature classification parameters are adjusted in real time, including adjusting the feature classification threshold, adding or deleting feature classification categories, or adjusting the feature classification weight.

[0028] Preferably, the step of filling the extracted timing features into the pre-built multiple simulation modules according to the updated data mapping rules includes:

[0029] Obtaining an updated data mapping rule, wherein the updated data mapping rule defines a corresponding relationship between timing features and simulation modules and a new standard for feature classification;

[0030] Traverse the extracted time series feature set, and classify and mark each time series feature according to the updated data mapping rules;

[0031] According to the classification and labeling results, the time series features are filled into the corresponding simulation modules according to the feature categories to which they belong, wherein each simulation module is pre-designed to receive time series features of a specific category as input.

[0032] Preferably, the time sequence determined based on the timing characteristics of each simulation module, and running the corresponding and pre-configured simulation modules on each distributed computing node of the pre-built distributed computing framework, includes:

[0033] Analyze the timing characteristics of each simulation module, determine their timestamps and the time dependencies between them, and build a time series model of the timing characteristics;

[0034] Determine the execution order and parallel execution strategy of each simulation module according to the time series model;

[0035] In a pre-built distributed computing framework, a corresponding distributed computing node is allocated to each simulation module, wherein the distributed computing framework supports data communication and synchronization mechanisms between nodes;

[0036] Deploy the configured simulation modules and their required data and parameters to the corresponding distributed computing nodes, ensuring that the simulation modules on each node can run independently and work in coordination with the simulation modules on other nodes;

[0037] Start the distributed computing framework, run the simulation modules in parallel or serially on each distributed computing node according to the determined time sequence and execution strategy, simulate the production status of the flexible brushless motor production line in real time, and maintain the consistency and accuracy of the simulation status through data communication and synchronization mechanisms between nodes.

[0038] The present invention also provides a flexible brushless motor production line digital twin system, which is used to execute a flexible brushless motor production line digital twin method, and the system comprises:

[0039] A feature extraction module, used to collect multi-source data sets from a flexible brushless motor production line and extract time series features from the multi-source data sets;

[0040] A strategy output module, used to analyze the extracted time series features using a recursive neural network and output a rule adjustment strategy, wherein the recursive neural network is used to identify the change characteristics of the time series features;

[0041] A rule updating module, used to update the preset data mapping rules in real time based on the rule adjustment strategy to obtain updated data mapping rules, wherein the preset data mapping rules are predefined feature classification rules;

[0042] A feature filling module, used to fill the extracted timing features into a plurality of pre-built simulation modules according to the updated data mapping rules, wherein the simulation modules are a plurality of independent but interrelated functional units pre-defined according to the production process of the flexible brushless motor production line;

[0043] The real-time simulation module is used to run the corresponding and pre-configured simulation modules on each distributed computing node of the pre-built distributed computing framework based on the time sequence determined by the timing characteristics of each simulation module, so as to simulate the production status of the flexible brushless motor production line in real time and realize the digital twin of the flexible brushless motor production line.

[0044] Compared with the related art, the digital twin method and system of a flexible brushless motor production line provided by the present invention have the following beneficial effects:

[0045] The present invention collects multi-source data sets and extracts time series features, uses recursive neural networks to analyze the time series features and outputs rule adjustment strategies, updates preset data mapping rules in real time, and fills the time series features into a pre-built simulation module.

[0046] At the same time, the method also adopts a distributed computing framework to realize the parallel operation and collaborative work of simulation modules, thereby simulating the production status of the flexible brushless motor production line in real time. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A flow chart of a digital twin method for a flexible brushless motor production line provided by the present invention;

[0048] Figure 2 A module structure diagram of a digital twin system for a flexible brushless motor production line provided by the present invention. DETAILED DESCRIPTION

[0049] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only the parts related to the present invention, rather than all structures, are shown in the accompanying drawings. In addition, the embodiments of the present invention and the features in the embodiments may be combined with each other without conflict.

[0050] It should also be noted that, for ease of description, only the parts related to the present invention, but not all of the contents, are shown in the accompanying drawings. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the operations (or steps) as sequential processes, many of the operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but it can also have additional steps not included in the accompanying drawings. The process can correspond to methods, functions, procedures, subroutines, subprograms, etc.

[0051] Embodiment 1

[0052] The present invention provides a digital twin method for a flexible brushless motor production line, referring to Figure 1 As shown, the method comprises the following steps:

[0053] S1: Collect multi-source data sets from a flexible brushless motor production line, and extract timing features from the multi-source data sets.

[0054] Specifically, step S1 includes the following steps:

[0055] S11: The operation data, equipment status data, quality control data and environmental parameter data on the flexible brushless motor production line are captured in real time through sensor networks and Internet of Things technologies to form the multi-source data set.

[0056] In this embodiment, in order to improve production efficiency and product quality, it is particularly important to monitor and analyze the operating status of the production line in real time. By deploying sensor networks and applying Internet of Things technology, the operating data, equipment status data, quality control data, and environmental parameter data on the flexible brushless motor production line can be captured in real time. These data together constitute a multi-source data set, which provides a basis for subsequent data processing and analysis. It ensures that the acquired data is highly timely and comprehensive, thereby providing the possibility for status monitoring and fault prediction of the production line.

[0057] S12: Preprocessing the multi-source data set, wherein the preprocessing includes removing noise data and irrelevant features.

[0058] In this embodiment, the original collected data often contains noise and irrelevant information, which will affect the accuracy and efficiency of subsequent data analysis. Therefore, it is necessary to preprocess the collected multi-source data sets, mainly to remove noise data and filter out features that are irrelevant to the target analysis. This process requires the use of filters and statistical methods to identify and remove outliers.

[0059] In addition, appropriate features will be selected according to analysis requirements, which helps to significantly improve the quality of the data, reduce unnecessary computational burden, and make subsequent feature extraction more accurate and efficient.

[0060] S13: extracting time series features from the preprocessed multi-source data set using a feature extraction algorithm, wherein the time series features include peak features, slope features, statistics features, spectrum density features, phase features, and transient behavior features.

[0061] In this embodiment, in order to have a deeper understanding of the operation mode and potential problems of the production line, it is necessary to extract key time series characteristics from a large amount of data. Specific feature extraction algorithms, including peak detection, slope calculation, statistical analysis (mean, variance, etc.), spectral density estimation, phase analysis, and transient behavior capture, are used to extract representative time series features from the data. The obtained time series features can accurately reflect the dynamic changes of the production line, laying a solid foundation for further analysis using recursive neural networks. At the same time, these features also provide strong data support for optimizing production processes and improving product quality.

[0062] S2: Analyze the extracted time series features using a recursive neural network and output a rule adjustment strategy, wherein the recursive neural network is used to identify change characteristics of the time series features.

[0063] Specifically, step S2 includes the following steps:

[0064] S21: The extracted time series features are used as input data and input into a pre-built and trained recursive neural network model.

[0065] In this embodiment, a pre-built and trained recursive neural network (RNN) model is prepared. The model is trained based on historical data to learn patterns in time series data. In actual operation, the time series features extracted in step S1 are fed into this model as input. These features include peak features, slope features, etc., which are transmitted to the RNN model through a data interface. The memory unit inside the RNN automatically adjusts its weights according to the input sequence in order to capture the dependencies in the time series. This mechanism can effectively identify and predict the changing trend of the production line status and provide a basis for subsequent optimization.

[0066] S22: The recursive neural network model captures the time dependency and long-term dependency in the time series features through its internal recursive structure to identify the change characteristics of the time series features, where the change characteristics include trend changes, periodic fluctuations and abnormal events.

[0067] In this embodiment, the recursive neural network model processes the input time series data through its internal recursive structure. The recursive neural network controls the information flow through a special gating mechanism, so that it can capture data dependencies over a long time span. During the operation of the model, the input at each moment is combined with the hidden state of the previous moment, and the output and new hidden state of the current moment are generated through a series of nonlinear transformations. These transformations allow the model to identify trend changes, periodic fluctuations and abnormal events in the data. For example, by analyzing the changes in data over a continuous time period, a trend of gradually decreasing production efficiency can be found or early signals of possible equipment failure can be detected, so that the production line can be monitored in real time and any factors that may affect production can be quickly responded to.

[0068] S23: Based on the identified change characteristics of the timing characteristics, the recursive neural network model outputs a rule adjustment strategy, and the rule adjustment strategy is used to guide the real-time update of the preset data mapping rules to optimize the simulation accuracy of the simulation module on the production status of the flexible brushless motor production line.

[0069] This step is to quantitatively analyze the identified changes in timing features. Then, based on the above analysis results, the recursive neural network model will automatically generate an initial rule adjustment strategy. The generated strategy will then be compared and verified with the preset standards to ensure that it meets the established quality and performance requirements.

[0070] Wherein, step S23 comprises the following steps:

[0071] S23a: Based on the change characteristics of the identified timing characteristics, determine the degree of impact and potential impact range on the production status of the flexible brushless motor production line.

[0072] In this embodiment, in order to ensure that an effective rule adjustment strategy is formulated, it is first necessary to accurately quantify the impact and potential impact range of the identified timing feature changes on the flexible brushless motor production line.

[0073] Specifically, statistical analysis and machine learning algorithms are used to evaluate these change characteristics. For example, the rate of change or fluctuation of data within a specific time period can be calculated to measure its impact on production efficiency.

[0074] Cluster analysis can also be used to identify specific parts of the production line that are affected. This process usually involves analyzing large amounts of historical data to find situations similar to the current change pattern and estimate the possible scope of the impact.

[0075] In this way, it is possible to accurately identify which production links are affected and the extent of the impact, thus providing a basis for the subsequent formulation of targeted adjustment strategies.

[0076] S23b: Based on the degree of impact and potential impact range, the recursive neural network model automatically calculates and generates an initial rule adjustment strategy.

[0077] In this embodiment, the recursive neural network model automatically calculates and generates an adjustment strategy based on the degree of impact and potential impact range determined in the previous step. This process relies on pre-set algorithmic logic that takes into account multiple factors, such as the current state of the production line, expected goals, and resource constraints.

[0078] For example, if a decrease in the efficiency of a certain production link is detected, it will be recommended to adjust the operating parameters of the link (such as speed, temperature, etc.) or reallocate resources.

[0079] In addition, the interaction between different adjustment measures is taken into account to ensure that the proposed strategy can comprehensively optimize the overall production performance. This makes the rule adjustment strategy not only targeted, but also dynamically adapts to the actual operation of the production line, improving the flexibility and efficiency of responding to changes.

[0080] S23c: Compare and verify the generated initial rule adjustment strategy with the preset strategy evaluation standard, and determine the final rule adjustment strategy based on the comparison and verification result.

[0081] In this embodiment, the generated initial rule adjustment strategy is compared and verified with preset strategy evaluation criteria. These evaluation criteria are usually established based on past experience, industry best practices or theoretical models, and they provide a benchmark for judging the effectiveness and feasibility of the strategy.

[0082] Next, evaluate the performance of the new strategy through simulation testing or historical data analysis to check whether it has achieved the expected results, such as improving production efficiency, reducing costs, or improving product quality.

[0083] Finally, based on the results of the comparative verification, necessary fine-tuning is performed on the strategy to ensure that it can not only meet the established quality and performance requirements but also be actually applied in the production line.

[0084] Through a strict verification process, unreasonable adjustment strategies can be effectively avoided from being adopted, ensuring the reliability and effectiveness of the measures taken, thereby improving the stability and competitiveness of the entire production system.

[0085] S3: Based on the rule adjustment strategy, the preset data mapping rule is updated in real time to obtain an updated data mapping rule, wherein the preset data mapping rule is a predefined feature classification rule.

[0086] Specifically, step S3 includes the following steps:

[0087] S31: receiving a rule adjustment strategy from a recursive neural network model, wherein the rule adjustment strategy adjusts a feature classification threshold, a feature classification category, or a feature classification weight in a preset data mapping rule.

[0088] In this embodiment, in order to ensure that the simulation module can accurately reflect the actual state of the flexible brushless motor production line, it is necessary to dynamically update the data mapping rules according to the rule adjustment strategy output by the recursive neural network model, specifically:

[0089] First, the rule adjustment strategy from the recurrent neural network model is received. This process involves collecting adjustment suggestions from the model output, which include specific modifications to feature classification thresholds, feature classification categories, or feature classification weights. For example, if it is found that temperature fluctuations in a certain production link have a significant impact on product quality, the adjustment strategy will recommend lowering the classification threshold of the feature to improve monitoring accuracy. This mechanism enables the data mapping rules to respond to changes on the production line in a timely manner, improving the accuracy and reliability of the simulation results.

[0090] S32: Analyze the preset data mapping rules to identify the feature classification parameters that need to be adjusted.

[0091] In this embodiment, it becomes crucial to parse the existing data mapping rules and identify the specific parameters that need to be adjusted. Specifically, it includes an in-depth analysis of the data mapping rules currently in use to determine which parts need to be modified according to the new rule adjustment strategy. This step involves algorithm analysis and logical reasoning to clarify which feature classification parameters (such as thresholds, categories, or weights) need to be adjusted. For example, if an increase in the frequency of certain types of faults is detected, it is necessary to add new feature classification categories or adjust the weights of existing categories. In this way, the parts that need to be optimized can be accurately located, laying the foundation for subsequent adjustments. This method ensures that only necessary adjustments are performed, avoids unnecessary complexity, and improves the flexibility and adaptability of the system.

[0092] S33: According to the received rule adjustment strategy, the identified feature classification parameters are adjusted in real time, including adjusting the feature classification threshold, adding or deleting the feature classification category, or adjusting the feature classification weight.

[0093] In this embodiment, the received rule adjustment strategy is applied according to the feature classification parameters that need to be adjusted identified in step S32. For example, for feature classification thresholds that need to be adjusted, they will be updated according to the newly set values; for feature classification categories that need to be added or deleted, the database or configuration file will be modified accordingly; and for the adjustment of feature classification weights, it involves retraining part of the machine learning model or adjusting algorithm parameters.

[0094] During this process, all changes need to be rigorously tested to ensure that they do not introduce new problems. Through such a dynamic adjustment mechanism, it is possible to quickly adapt to changes in the production line and maintain the efficiency and accuracy of the data processing process, thereby supporting more accurate simulation and prediction of production status. In addition, this also enhances self-optimization capabilities and helps to continuously improve production efficiency and product quality.

[0095] S4: According to the updated data mapping rules, the extracted timing features are correspondingly filled into a plurality of pre-built simulation modules, wherein the simulation modules are a plurality of independent but interrelated functional units pre-defined according to the production process of the flexible brushless motor production line.

[0096] Specifically, step S4 includes the following steps:

[0097] S41: Obtain updated data mapping rules, wherein the updated data mapping rules define the corresponding relationship between timing features and simulation modules and a new standard for feature classification.

[0098] In this embodiment, in order to ensure that the simulation module can accurately reflect the actual operating status of the production line, the data mapping rules must be adjusted according to the latest production situation. Specifically, the updated data mapping rules are obtained, which not only define the correspondence between the timing features and the simulation modules, but also include new standards for feature classification. This process involves receiving the latest rule adjustment strategy from the recursive neural network model and integrating it into the existing data mapping rules. In this way, it can be ensured that the data received by each simulation module is the most in line with the current production conditions, thereby improving the accuracy of the simulation. The technical effect is that this dynamic adjustment mechanism enables the simulation module to more accurately simulate the actual operating conditions of the production line, providing reliable data support for optimizing the production process.

[0099] S42: traverse the extracted time series feature set, and classify and mark each time series feature according to the updated data mapping rule.

[0100] In this embodiment, considering that different types of timing features have different meanings for different simulation modules, it is necessary to carefully classify and mark the timing features. Specifically, the extracted timing feature set is traversed, and each timing feature is classified and marked according to the updated data mapping rules. For example, some features may be marked as features related to the health status of the equipment, while others may be related to product quality control. In this way, it can be ensured that each timing feature can be correctly assigned to the simulation module that best suits it. The efficiency and accuracy of data processing are greatly improved, so that the simulation module can make decisions based on more accurate data, thereby improving the performance of the entire system.

[0101] S43: Based on the classification and labeling results, the timing features are filled into corresponding simulation modules according to the feature categories to which they belong, wherein each simulation module is pre-designed to receive timing features of a specific category as input.

[0102] In this embodiment, according to the classification and labeling results obtained in the previous step, the time series features are filled into the corresponding simulation modules according to the feature categories to which they belong. Each simulation module is pre-designed to receive a specific category of time series features as input, which means that they can specifically process a specific type of data related to it. For example, the simulation module responsible for monitoring the health status of the device will only receive those time series features that are closely related to the device status.

[0103] S5: Based on the time sequence determined by the timing characteristics of each simulation module, the corresponding and pre-configured simulation modules are run on each distributed computing node of the pre-built distributed computing framework to simulate the production status of the flexible brushless motor production line in real time and realize the digital twin of the flexible brushless motor production line.

[0104] Specifically, step S5 includes the following steps:

[0105] S51: Analyze the timing characteristics of each simulation module, determine their timestamps and the time dependencies between them, and build a time series model of the timing characteristics.

[0106] In this embodiment, the timing features in each simulation module are first analyzed in depth, including determining the timestamp of each feature and its time dependency with other features. This process usually involves the application of time series analysis techniques, including but not limited to autocorrelation function (ACF) and partial autocorrelation function (PACF), in order to build an accurate time series model. This model not only helps to understand the trend of each feature over time, but also lays the foundation for the execution order and strategy formulation of the simulation modules in the subsequent steps. In this way, the key information of the dynamic changes of the production line can be captured more accurately, providing strong data support for subsequent simulations.

[0107] S52: Determine the execution order and parallel execution strategy of each simulation module according to the time series model.

[0108] In this embodiment, according to the time series model constructed in step S51, the interactions and dependencies between the simulation modules are further analyzed to determine their execution order and which modules can be executed in parallel. For example, if the input of some modules depends on the output of other modules, these modules should be executed in sequence; for those modules that are independent of each other, parallel execution can be considered to save time. This process may also need to be combined with scheduling algorithms, such as priority scheduling or shortest job first, to optimize the performance of the entire system. By reasonably arranging the execution order and strategy of the simulation modules, not only the efficiency of the simulation can be improved, but also the accuracy and reliability of the simulation results can be ensured.

[0109] S53: In a pre-built distributed computing framework, a corresponding distributed computing node is allocated to each simulation module, wherein the distributed computing framework supports data communication and synchronization mechanisms between nodes.

[0110] In this embodiment, in a complex production environment, in order to efficiently handle a large number of simulation tasks, it is necessary to utilize the advantages of a distributed computing framework to allocate resources. Specifically, based on the execution order and parallel strategy determined in the previous step, a corresponding distributed computing node is allocated to each simulation module in a pre-built distributed computing framework. This process takes into account factors such as load balancing and communication delay between nodes to ensure that each module can run in its optimal environment. In addition, since the distributed computing framework supports data communication and synchronization mechanisms between nodes, it is also necessary to configure appropriate protocols and interfaces to facilitate data exchange and state synchronization. This resource allocation method not only improves the scalability and flexibility of the simulation, but also enhances the fault tolerance of the system, so that even if some nodes fail, the continuity of the overall simulation process can be maintained.

[0111] S54: deploy the configured simulation modules and their required data and parameters to the corresponding distributed computing nodes, ensuring that the simulation modules on each node can run independently and work in coordination with the simulation modules on other nodes.

[0112] In this embodiment, in order to ensure that each simulation module can run smoothly on the computing node assigned to it and can work with other modules. Specifically, the previously designed simulation module and all the data and parameters required by it are deployed to the corresponding distributed computing nodes. This includes setting the correct environment variables, installing necessary software dependencies, and uploading related data files. At the same time, it is also necessary to configure network connections and security settings to ensure that simulation modules on different nodes can communicate seamlessly. In this way, it is ensured that all simulation modules can run independently and efficiently according to the predetermined plan, and it also promotes collaboration between modules to jointly complete the real-time simulation of the production line status.

[0113] S55: Start the distributed computing framework, run the simulation modules in parallel or serially on each distributed computing node according to the determined time sequence and execution strategy, simulate the production status of the flexible brushless motor production line in real time, and maintain the consistency and accuracy of the simulation status through data communication and synchronization mechanism between nodes.

[0114] In this embodiment, according to the previously determined time sequence and execution strategy, the distributed computing framework is started, and the simulation modules are started to run on each computing node. In this process, the status of each node is continuously monitored to ensure that they are executed as expected. At the same time, the data communication and synchronization mechanism between nodes are used to maintain the consistency and accuracy of all simulation modules. Once a problem is found, timely measures are taken to adjust it. This method not only realizes the real-time simulation of the production status of the flexible brushless motor production line, but also ensures the reliability and effectiveness of the simulation through continuous monitoring and adjustment, providing a solid foundation for the optimization of the production line.

[0115] Embodiment 2

[0116] The present invention also provides a flexible brushless motor production line digital twin system for executing a flexible brushless motor production line digital twin method, referring to Figure 2 As shown, the system comprises:

[0117] The feature extraction module 100 is used to collect multi-source data sets from a flexible brushless motor production line and extract time series features from the multi-source data sets.

[0118] The strategy output module 200 is used to analyze the extracted time series features using a recursive neural network and output a rule adjustment strategy, wherein the recursive neural network is used to identify the change characteristics of the time series features.

[0119] The rule updating module 300 is used to update the preset data mapping rules in real time based on the rule adjustment strategy to obtain updated data mapping rules, wherein the preset data mapping rules are predefined feature classification rules.

[0120] The feature filling module 400 is used to fill the extracted timing features into a plurality of pre-built simulation modules according to the updated data mapping rules, wherein the simulation modules are a plurality of independent but interrelated functional units pre-defined according to the production process of the flexible brushless motor production line.

[0121] The real-time simulation module 500 is used to run the corresponding and pre-configured simulation modules on each distributed computing node of the pre-built distributed computing framework based on the time sequence determined by the timing characteristics of each simulation module, so as to simulate the production status of the flexible brushless motor production line in real time and realize the digital twin of the flexible brushless motor production line.

[0122] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0123] A person skilled in the art may understand that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, the storage medium including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically-erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0124] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

Claims

1. A digital twin method for a flexible brushless motor production line, characterized in that: The method comprises the following steps: Collecting a multi-source data set from a flexible brushless motor production line, and extracting time series features from the multi-source data set; Using a recursive neural network to analyze the extracted time series features and output a rule adjustment strategy, wherein the recursive neural network is used to identify change characteristics of the time series features; Based on the rule adjustment strategy, the preset data mapping rule is updated in real time to obtain an updated data mapping rule, wherein the preset data mapping rule is a predefined feature classification rule; This step specifically includes: Receiving a rule adjustment strategy from a recursive neural network model, wherein the rule adjustment strategy adjusts a feature classification threshold, a feature classification category, or a feature classification weight in a preset data mapping rule; Parse the preset data mapping rules and identify the feature classification parameters that need to be adjusted; According to the received rule adjustment strategy, the identified feature classification parameters are adjusted in real time, including adjusting the feature classification threshold, adding or deleting feature classification categories, or adjusting the feature classification weight; According to the updated data mapping rules, the extracted timing features are correspondingly filled into a plurality of pre-built simulation modules, wherein the simulation modules are a plurality of independent but interrelated functional units pre-defined according to the production process of the flexible brushless motor production line; Based on the time sequence determined by the timing characteristics of each simulation module, the corresponding and pre-configured simulation modules are run on each distributed computing node of the pre-built distributed computing framework to simulate the production status of the flexible brushless motor production line in real time and realize the digital twin of the flexible brushless motor production line.

2. A digital twin method for a flexible brushless motor production line according to claim 1, characterized in that: The method collects a multi-source data set from a flexible brushless motor production line and extracts time series features from the multi-source data set, including: Capturing the operation data, equipment status data, quality control data and environmental parameter data on the flexible brushless motor production line in real time through sensor networks and Internet of Things technologies to form the multi-source data set; Preprocessing the multi-source data set, wherein the preprocessing includes removing noise data and irrelevant features; A feature extraction algorithm is used to extract time series features from the preprocessed multi-source data set, wherein the time series features include peak features, slope features, statistics features, spectrum density features, phase features and transient behavior features.

3. A digital twin method for a flexible brushless motor production line according to claim 2, characterized in that: The recursive neural network is used to analyze the extracted time series features and output a rule adjustment strategy, wherein the recursive neural network is used to identify the change characteristics of the time series features, including: The extracted time series features are used as input data into a pre-built and trained recurrent neural network model; The recursive neural network model captures the time dependency and long-term dependency in the time series features through its internal recursive structure to identify the change characteristics of the time series features, wherein the change characteristics include trend changes, periodic fluctuations and abnormal events; Based on the identified change characteristics of the timing characteristics, the recursive neural network model outputs a rule adjustment strategy, and the rule adjustment strategy is used to guide the real-time update of the preset data mapping rules to optimize the simulation accuracy of the simulation module on the production status of the flexible brushless motor production line.

4. A digital twin method for a flexible brushless motor production line according to claim 3, characterized in that: The recursive neural network model output rule adjustment strategy based on the identified change characteristics of the time series characteristics includes: Based on the change characteristics of the identified timing characteristics, determine the degree of impact and potential impact range on the production status of the flexible brushless motor production line; According to the degree of impact and potential impact range, the recursive neural network model automatically calculates and generates an initial rule adjustment strategy; The generated initial rule adjustment strategy is compared and verified with the preset strategy evaluation criteria, and the final rule adjustment strategy is determined based on the comparison and verification results.

5. A digital twin method for a flexible brushless motor production line according to claim 4, characterized in that: The extracted timing features are correspondingly filled into a plurality of pre-built simulation modules according to the updated data mapping rules, including: Obtaining an updated data mapping rule, wherein the updated data mapping rule defines a corresponding relationship between timing features and simulation modules and a new standard for feature classification; Traverse the extracted time series feature set, and classify and mark each time series feature according to the updated data mapping rules; According to the classification and labeling results, the time series features are filled into the corresponding simulation modules according to the feature categories to which they belong, wherein each simulation module is pre-designed to receive time series features of a specific category as input.

6. A digital twin method for a flexible brushless motor production line according to claim 5, characterized in that: The time sequence determined based on the timing characteristics of each simulation module, and running the corresponding and pre-configured simulation modules on each distributed computing node of the pre-built distributed computing framework, includes: Analyze the timing characteristics of each simulation module, determine their timestamps and the time dependencies between them, and build a time series model of the timing characteristics; Determine the execution order and parallel execution strategy of each simulation module according to the time series model; In a pre-built distributed computing framework, a corresponding distributed computing node is allocated to each simulation module, wherein the distributed computing framework supports data communication and synchronization mechanisms between nodes; Deploy the configured simulation modules and their required data and parameters to the corresponding distributed computing nodes, ensuring that the simulation modules on each node can run independently and work in coordination with the simulation modules on other nodes; Start the distributed computing framework, run the simulation modules in parallel or serially on each distributed computing node according to the determined time sequence and execution strategy, simulate the production status of the flexible brushless motor production line in real time, and maintain the consistency and accuracy of the simulation status through data communication and synchronization mechanisms between nodes.

7. A flexible brushless motor production line digital twin system, used to execute a flexible brushless motor production line digital twin method as claimed in any one of claims 1 to 6, characterized in that: The system comprises: A feature extraction module, used to collect multi-source data sets from a flexible brushless motor production line and extract time series features from the multi-source data sets; A strategy output module, used to analyze the extracted time series features using a recursive neural network and output a rule adjustment strategy, wherein the recursive neural network is used to identify the change characteristics of the time series features; A rule updating module, used to update the preset data mapping rules in real time based on the rule adjustment strategy to obtain updated data mapping rules, wherein the preset data mapping rules are predefined feature classification rules; A feature filling module, used to fill the extracted timing features into a plurality of pre-built simulation modules according to the updated data mapping rules, wherein the simulation modules are a plurality of independent but interrelated functional units pre-defined according to the production process of the flexible brushless motor production line; The real-time simulation module is used to run the corresponding and pre-configured simulation modules on each distributed computing node of the pre-built distributed computing framework based on the time sequence determined by the timing characteristics of each simulation module, so as to simulate the production status of the flexible brushless motor production line in real time and realize the digital twin of the flexible brushless motor production line.

Images