System and method for improving adaptability of manufacturing and processing environment of aluminum alloy parts based on new energy

Abstract

The invention discloses an adaptability improving system and method based on a new energy aluminum alloy part manufacturing and machining environment, and relates to the technical field of new energy aluminum alloy part manufacturing. The system comprises the following components: a data acquisition module, a digital twin model construction module, a data transmission and processing module, a quality tracing and analysis module and a decision and optimization module. According to the new energy aluminum alloy part manufacturing and processing full-process digital twinborn model, the processing equipment operation parameters, the part processing process data and the processing environment data are comprehensively collected, and the new energy aluminum alloy part manufacturing and processing full-process digital twinborn model is constructed, so that the processing process is accurately simulated and monitored in real time; the system can quickly position environmental factors and processing links which cause quality problems, and generates a processing technology optimization scheme in combination with a decision and optimization module, and the adaptive capacity of new energy aluminum alloy part manufacturing and processing to environmental changes is remarkably improved through a series of measures.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy aluminum alloy component manufacturing, and specifically to a system and method for improving the adaptability of the processing environment for manufacturing new energy aluminum alloy components. Background Art

[0002] With the rapid development of the new energy industry, aluminum alloy parts, as key components of new energy equipment, have a vital impact on the performance, safety and reliability of the entire equipment. However, the manufacturing process of new energy aluminum alloy parts is complex and changeable, involving a variety of processing equipment, process flows and environmental factors. The interaction and influence between these factors make it particularly difficult to control the processing quality, especially under different processing environmental conditions, such as temperature, humidity, dust concentration and electromagnetic interference, which will have a significant impact on the processing accuracy and surface quality of parts.

[0003] In the traditional manufacturing process of aluminum alloy parts, the monitoring and control of environmental factors often rely on manual experience and simple sensor equipment. This method has many limitations: first, manual monitoring is difficult to achieve real-time, comprehensive and accurate, and it is easy to miss key information; second, simple sensor equipment can often only monitor a single environmental parameter and cannot comprehensively consider the comprehensive impact of multiple environmental factors on processing quality; third, the lack of effective data processing and analysis methods makes it difficult to accurately judge the specific impact of data on processing quality even if it is collected; finally, traditional technologies often lack intelligent decision-making and optimization mechanisms, and cannot adjust processing parameters in time according to real-time processing conditions and environmental changes, thereby limiting the improvement of processing quality and efficiency.

[0004] In summary, the traditional manufacturing and processing technology of new energy aluminum alloy parts has significant deficiencies in environmental adaptability, and it is difficult to meet the current new energy industry's demand for high-precision, high-efficiency and high-reliability parts. Therefore, it is particularly important to develop a system and method for improving the environmental adaptability of the manufacturing and processing of new energy aluminum alloy parts. Summary of the Invention

[0005] The purpose of the present invention is to make up for the shortcomings of the existing technology and provide a system and method for improving the adaptability of the manufacturing processing environment of new energy aluminum alloy parts. It can realize intelligent decision-making and optimization by real-time, comprehensive and accurate monitoring of processing environment data, combined with advanced data processing and analysis technology, thereby significantly improving the processing quality and efficiency of parts.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: On the one hand, based on the system for improving the adaptability of the manufacturing and processing environment of new energy aluminum alloy parts, the system includes the following components: a data acquisition module, a digital twin model construction module, a data transmission and processing module, a quality traceability and analysis module, and a decision-making and optimization module;

[0007] The data acquisition module is used to deploy various sensors at the manufacturing site of new energy aluminum alloy parts to collect processing equipment operating parameters, parts processing process data and processing environment data;

[0008] The digital twin model construction module is used to construct a digital twin model of the entire manufacturing process of new energy aluminum alloy parts based on the collected data;

[0009] The data transmission and processing module transmits the collected data to the digital twin model construction module, cleans, filters and pre-processes the data, and then updates the data to the digital twin model synchronously;

[0010] The quality traceability and analysis module compares the virtual quality data of parts in the digital twin model with the actual inspection quality data, and locates the environmental factors and processing links that cause quality problems in combination with the processing environment data;

[0011] The decision-making and optimization module generates a processing technology optimization plan based on the quality traceability and analysis results, combined with historical processing data and process experience, and feeds back to the processing equipment and processing environment control system.

[0012] Furthermore, in the data acquisition module, the deployment of sensors adopts a hierarchical and partitioned dynamic optimization strategy. At the processing equipment layer, vibration sensors and displacement sensors are deployed at the spindle and guide rail parts according to the stress and wear conditions of the key components of the equipment to collect the equipment operation vibration frequency and displacement change parameters. At the parts processing process layer, temperature sensors and pressure sensors are set at the processing tools and workpiece fixtures to obtain cutting temperature and cutting force data in real time. At the processing environment layer, temperature and humidity sensors, dust concentration sensors and electromagnetic interference sensors are dispersedly deployed in different areas according to the workshop space layout and environmental characteristics. The determination of the sensor deployment location and number is optimized based on the genetic particle swarm hybrid algorithm. This algorithm combines the global search capability of the genetic algorithm and the local search advantage of the particle swarm algorithm, with minimizing the part quality fluctuation as the objective function. Where X is the sensor deployment plan vector, n is the number of quality indicators, and w i For the weight of each quality indicator, 10 industry experts were invited to score the importance of each quality indicator through the hierarchical analysis method, and a judgment matrix was constructed and calculated. iThe optimal sensor deployment scheme is determined through iterative optimization of the algorithm for the fluctuation value of the i-th quality indicator to ensure the comprehensiveness and effectiveness of the collected data, provide an accurate data basis for the subsequent digital twin model construction and quality traceability, and enhance the system's perception of the processing environment.

[0013] Furthermore, in the digital twin model construction module, a multi-scale fusion modeling method is adopted;

[0014] At the microscale, molecular dynamics simulation methods are used to construct atomic-level structural models of aluminum alloys, simulate interatomic interactions and crystal structure changes, and analyze the impact of material microstructural evolution on component performance during processing;

[0015] At the mesoscopic scale, based on the finite element method, the local structure and key parts of the components are finely modeled to study their mechanical properties and deformation behavior;

[0016] At the macro scale, the entire manufacturing system model is established through 3D modeling software, including processing equipment, process flow, logistics and transportation, etc. The models of each scale realize information interaction and collaboration through data interfaces. During the model construction process, the model parameters are determined by the adaptive Kalman filter algorithm. Based on the traditional Kalman filter, the algorithm introduces the adaptive factor λ with a value range of [0.1, 0.9]. It is adjusted in real time according to the dynamic changes of the data. The state estimation equation is: in is the optimal state estimate at time k, is the predicted state at time k based on time k-1, K k is the Kalman gain, Z k is the measured value, H k is the measurement matrix, and the adaptive factor λ is based on the error between the measured value and the predicted value. Make adjustments when |e k When |>θ, λ=λ+0.1, which speeds up the update of model parameters. θ is the set threshold. It is found through statistics of a large number of experimental data that when |e k When |≤θ, λ=λ-0.1, which improves the stability of parameter estimation. Through multi-scale fusion modeling and adaptive parameter adjustment, the digital twin model can more accurately reflect the actual state of the physical entity and enhance the system's simulation and analysis capabilities of the processing process and environment.

[0017] Furthermore, in the data transmission and processing module, data cleaning adopts an abnormal data identification method based on fuzzy rough sets. First, the collected data is divided into multiple attribute sets according to different types. For each attribute set, a fuzzy rough set model is constructed and a fuzzy membership function is defined. Where a and b are parameters, and the historical normal data are fitted and determined by the particle swarm optimization algorithm so that the membership function can accurately reflect the normal distribution range of the data, and then the upper approximate set of the data is calculated. and lower approximation set R (X), abnormal data is satisfied The data is screened using a feature selection algorithm based on information entropy to calculate the information entropy of each data feature. where p(x i ) is the feature x i The probability of occurrence is set, the information entropy threshold τ is determined by cross-validation method, the data features with information entropy greater than τ are retained, redundant data are removed, and the data after cleaning and screening are normalized Perform preprocessing to ensure the accuracy and consistency of data, improve data transmission efficiency and the operating accuracy of digital twin models, and reduce quality traceability errors caused by data problems.

[0018] Furthermore, in the quality tracing and analysis module, a quality problem location model based on an improved convolutional neural network-long short-term memory network is adopted. In the CNN part, multiple convolution layers and pooling layers are designed, and the convolution layer adopts a custom convolution kernel K ij , its parameters are optimized by genetic algorithm to maximize the extraction of data features. For the input processing data and environmental data X, after convolution operation Extract local features and then reduce the dimension through the pooling layer. The LSTM part receives the features extracted by CNN and passes the forget gate f t =σ(W f [h t-1 ,x t ]+b f ), input gate i t =σ(W i [h t-1 ,x t ]+b i ), cell state update C t =f t C t-1 +i t tanh(W C [h t-1 ,x t ]+b C ) and output gate o t =σ(W o [h t-1 ,x t ]+b o ) processes the data, where σ is the activation function, W f 、W i、W C 、W o is the weight matrix, b f 、b i 、b C 、b o These parameters are trained through the back propagation algorithm and the adaptive moment estimation optimization algorithm. During the training process, the accuracy of quality problem location is used as the objective function. The parameters are adjusted so that the model can accurately identify the environmental factors and processing links that cause quality problems. At the same time, the attention mechanism is introduced to calculate the attention weight of each data feature on the quality problem. where e t =v T tanh(W a h t +b a ), v, W a 、b a The parameters are determined through training, focusing on key data features according to the attention weight, improving the accuracy and efficiency of quality traceability, and providing a reliable basis for decision-making and optimization.

[0019] Furthermore, in the decision-making and optimization module, the processing technology optimization scheme is generated by an optimization method based on a multi-objective bat algorithm, with the parts processing quality, processing efficiency and processing cost as multi-objective functions, respectively expressed as f1(x), f2(x), f3(x), where x is the processing technology parameter vector, and a comprehensive objective function F(x) = ω1f1(x) + ω2f2(x) + ω3f3(x) is established. ω1, ω2, ω3 are the weights of each objective, which are determined by the grey correlation analysis method. An evaluation team is invited to form an enterprise production management personnel, process engineers and quality experts to score the parts quality, processing efficiency and processing cost under different processing technologies, construct a grey correlation matrix, calculate the correlation degree of each objective, and then determine the weight. In the bat algorithm, the position of the individual bat represents the processing technology parameter scheme, and by adjusting the frequency f of the bat i =f min +(f max -f min )β, β is a random number with a value range of [0, 1], speed and location Search for x * In order to obtain the current optimal solution, an elite strategy is introduced during the search process to retain a certain number of individuals near the historical optimal solution, accelerate the convergence of the algorithm, and obtain the Pareto frontier solution set through algorithm iteration. According to the actual needs and production conditions of the enterprise, the optimal processing parameter scheme is selected from the solution set to achieve the optimization of the processing technology, effectively improving the adaptability of the processing environment for the manufacturing of new energy aluminum alloy parts and the economic benefits of the enterprise.

[0020] Furthermore, the system also includes an environmental intelligent control submodule, which is connected to the data acquisition module, the quality tracing and analysis module, and the decision and optimization module. The environmental intelligent control submodule intelligently controls the processing environment based on the environmental factors affecting the quality determined by the quality tracing and analysis module and the optimization plan generated by the decision and optimization module. When it is detected that the temperature is too high and affects the processing accuracy of the parts, the environmental intelligent control submodule adjusts the temperature by controlling the workshop air-conditioning system. Its control strategy adopts a dynamic temperature control algorithm based on reinforcement learning to construct a Markov decision process. The state space S is the temperature value of different areas of the workshop, and the action space A is the cooling power adjustment gear of the air-conditioning system. The reward function R is set according to the degree of influence of temperature changes on the quality of the parts. When the temperature is close to the optimal processing temperature range, a positive reward is given, and a negative reward is given when it deviates. Through continuous interactive learning between the intelligent agent and the environment, the optimal strategy π is obtained, so that the air-conditioning system adjusts the temperature according to the optimal strategy. In the case of excessively high dust concentration, the environmental intelligent control submodule controls the operation of the dust removal equipment and adopts a fuzzy PID control algorithm to adjust the power of the dust removal equipment. According to the deviation e between the dust concentration detection value and the set value and the deviation change rate ec, the PID parameter K is output through the fuzzy controller. p , K i , K d The adjustment amount can accurately control the power of the dust removal equipment, maintain the stability of the processing environment, reduce the adverse effects of environmental factors on the quality of parts, and further improve the system's environmental adaptability.

[0021] Furthermore, the system also features remote collaborative operation and maintenance capabilities, enabled by 5G networks and blockchain technology. In terms of data transmission, the high bandwidth and low latency characteristics of the 5G network are leveraged to rapidly transmit real-time data from the processing site to the remote operation and maintenance center. Blockchain technology is also used to store and manage data. Each data block contains a data hash value, a timestamp, and the hash value of the previous data block, ensuring the data's immutability and traceability. Experts at the remote operation and maintenance center can use the digital twin model to view the processing site's conditions in real time and remotely evaluate and provide guidance on quality traceability and analysis results. When complex quality issues arise, experts can initiate multi-party collaborative discussions on the blockchain platform. The opinions and suggestions of each participant are recorded on the blockchain in the form of smart contracts, ensuring the fairness and traceability of the discussion process. Based on the discussion results, experts at the remote operation and maintenance center use the decision-making and optimization modules to generate optimization solutions. These solutions are then securely and accurately transmitted to the control system at the processing site via blockchain smart contracts, enabling remote collaborative operation and maintenance. This function overcomes geographical restrictions, integrates resources from multiple parties, improves the system's efficiency in resolving quality issues and optimizing processing techniques, and enhances the system's adaptability and reliability in different application scenarios.

[0022] Furthermore, the system also includes a knowledge graph construction and application module, which is associated with the digital twin model construction module, the quality traceability and analysis module, and the decision and optimization module. The knowledge graph construction module collects multi-source knowledge such as process knowledge, material property knowledge, equipment maintenance knowledge, etc. in the field of aluminum alloy parts manufacturing and processing, extracts, classifies and semantically analyzes the knowledge through natural language processing technology, and constructs a knowledge graph based on entities and relationships. In the quality traceability process, the quality traceability and analysis module can assist in analyzing the potential relationship between environmental factors and quality problems based on the knowledge in the knowledge graph. In the decision and optimization stage, the decision and optimization module uses the knowledge in the knowledge graph and combines the analysis results of the digital twin model to generate a more optimized processing technology plan. According to the equipment maintenance knowledge and processing technology knowledge in the knowledge graph, the equipment maintenance plan and process parameter adjustment plan are determined to achieve efficient use of knowledge and intelligent decision-making of the system, further improving the system's ability and effect in improving the adaptability of the manufacturing and processing environment of new energy aluminum alloy parts.

[0023] On the other hand, a method for improving the adaptability of the manufacturing and processing environment of new energy aluminum alloy parts is characterized in that the method comprises the following specific steps:

[0024] S1. Data collection step: During the manufacturing process of new energy aluminum alloy parts, the data collection module is used to collect the operating parameters of the processing equipment, the parts processing process data and the processing environment data in real time;

[0025] S2. Digital twin model construction steps: Using the data acquired by the data acquisition module, a digital twin model of the entire manufacturing process of new energy aluminum alloy parts is constructed in the digital twin model construction module. During the model construction process, the interrelationships and influencing factors between each link are fully considered to ensure the accuracy and completeness of the model;

[0026] S3, data transmission and synchronization step: The data transmission and processing module processes the collected data and transmits it to the digital twin model construction module in real time, updates various parameters in the digital twin model, and realizes dynamic synchronization between the physical entity and the virtual model;

[0027] S4. Quality traceability and analysis step: Periodically or when quality issues are discovered, the quality traceability and analysis module compares the virtual quality data of the components in the digital twin model with the actual quality data. When quality deviations occur, the module combines the processing environment data, applies data analysis algorithms and machine learning models, analyzes the impact of environmental factors on the quality issues, and identifies the specific environmental factors and processing links that lead to the quality issues.

[0028] S5. Decision-making and optimization steps: The decision-making and optimization module uses the optimization algorithm to generate a processing optimization plan based on the results of the quality traceability and analysis module, combined with historical processing data and process experience, and sends the plan to the processing equipment and processing environment control system to adjust the processing process and improve the adaptability of the parts manufacturing processing environment.

[0029] Compared with the existing technology, the system and method for improving the adaptability of the manufacturing and processing environment of new energy aluminum alloy parts have the following beneficial effects:

[0030] 1. This system comprehensively collects processing equipment operating parameters, component processing data, and processing environment data, and constructs a digital twin model of the entire manufacturing process of new energy aluminum alloy components, achieving accurate simulation and real-time monitoring of the processing process. Through the quality traceability and analysis module, the system can quickly locate the environmental factors and processing links that lead to quality problems, and combine the decision-making and optimization module to generate processing technology optimization solutions. This series of measures significantly improves the adaptability of new energy aluminum alloy component manufacturing to environmental changes, ensuring the processing quality and stability of components under different environmental conditions.

[0031] 2. This system has the function of remote collaborative operation and maintenance. It uses the high bandwidth and low latency characteristics of the 5G network to quickly transmit real-time data from the processing site to the remote operation and maintenance center. At the same time, blockchain technology is used to store and manage data to ensure that the data cannot be tampered with and is traceable. Experts at the remote operation and maintenance center can view the processing site conditions in real time through the digital twin model, and remotely evaluate and guide quality traceability and analysis results. When complex quality problems arise, experts can initiate multi-party collaborative discussions on the blockchain platform and generate optimization plans. In addition, the system also includes a knowledge graph construction and application module, which provides rich knowledge support for decision-making and optimization. These functions jointly realize remote collaborative operation and maintenance and intelligent decision-making, greatly improving production efficiency and economic benefits.

[0032] Other advantages, objects and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be learned from the practice of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0034] Figure 1 To provide a system process operation diagram for improving the adaptability of the processing environment for manufacturing aluminum alloy parts based on new energy;

[0035] Figure 2 This is a process operation diagram for the method of improving the adaptability of the processing environment in the manufacturing of new energy aluminum alloy parts. DETAILED DESCRIPTION

[0036] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0037] Example 1

[0038] This embodiment describes the application of a system for improving the adaptability of the manufacturing and processing environment for new energy aluminum alloy parts in a modern automobile aluminum alloy wheel hub production workshop, effectively ensuring high-quality production of the hub hub.

[0039] At the processing equipment level, technicians precisely deploy vibration sensors and displacement sensors on the spindle and guide rails of the wheel hub processing lathe, which are prone to failure and have a great impact on processing accuracy, based on the stress and wear conditions of the key components of the equipment. The deployment location and number of sensors are not determined arbitrarily, but are determined by using a genetic particle swarm hybrid algorithm with the objective function of minimizing wheel hub quality fluctuations. Optimize, where X represents the sensor deployment plan vector, covering the specific location information of each sensor, n is the number of quality indicators, such as the roundness of the wheel hub, dynamic balance performance and other indicators, w i is the weight of each quality index, reflecting the importance of different quality indicators to the overall quality of the hub, q i is the fluctuation value of the i-th quality indicator. At the component processing layer, temperature sensors and pressure sensors are cleverly set at the processing tools and workpiece fixtures. During the cutting process, excessively high tool temperature will accelerate tool wear and affect processing accuracy. Unstable pressure may cause defects on the hub surface. These sensors can obtain cutting temperature and cutting force data in real time. At the processing environment layer, considering the workshop space layout and environmental characteristics, temperature and humidity sensors, dust concentration sensors and electromagnetic interference sensors are dispersedly deployed in different areas. For example, in the area near the hub grinding, where the dust concentration is high, dust concentration sensors are deployed as a priority. Electromagnetic interference sensors are deployed in areas where electrical equipment is concentrated to ensure comprehensive and accurate collection of processing environment data.

[0040] A multi-scale fusion modeling method is used to construct a digital twin model of the entire wheel hub manufacturing process from the micro, meso and macro levels. At the micro scale, the molecular dynamics simulation method is used to construct an atomic-level structural model of aluminum alloy. In this microscopic world, the interaction between atoms and the changes in crystal structure are simulated, and the influence of the evolution of the material microstructure on the performance of the wheel hub during the processing is analyzed. For example, how the change in the atomic arrangement structure of aluminum alloy affects the strength and toughness of the wheel hub during high-speed cutting is studied. At the meso scale, the finite element method is used to finely model the local structures and key parts of the wheel hub, such as the bolt holes and rims, and their mechanical properties and deformation behavior are deeply studied, and the stress distribution and deformation trend of these parts under different working conditions are predicted. At the macro scale, a model of the entire manufacturing and processing system is established through 3D modeling software, covering various links such as processing equipment, process flow, logistics and transportation. Information interaction and collaboration are achieved between models of different scales through data interfaces. During the model construction process, the adaptive Kalman filter algorithm is used to determine the model parameters, and the state estimation equation is X k|k =X k|k-1 +K k (Z k -H k X k|k-1 ) to ensure that the model can be dynamically adjusted according to the real-time collected data to accurately reflect the actual production process.

[0041] The collected data are of various types and poor quality. First, they are cleaned by the abnormal data identification method based on fuzzy rough set. The technicians divide the collected data into multiple attribute sets according to different types. For each attribute set, a fuzzy rough set model is constructed and the fuzzy membership function is defined. (a and b are parameters), and then calculate the upper approximation set of the data and lower approximation set R (X), abnormal data is satisfied Then, we use the feature selection algorithm based on information entropy to filter the data and calculate the information entropy of each data feature. is feature x i The probability of occurrence) is set, the information entropy threshold τ is set, the data features with information entropy greater than τ are retained, redundant data are removed, and the data after cleaning and screening are normalized. Perform preprocessing to make it more in line with model processing requirements.

[0042] Using the quality problem location model based on the improved convolutional neural network-long short-term memory network, in the CNN part, multi-layer convolution layer and pooling layer are designed. The convolution layer uses a custom convolution kernel K optimized by genetic algorithm. ij, in order to maximize the extraction of data features, for the input processing data and environmental data X, after convolution operation Extract local features and then reduce the dimension through the pooling layer. The LSTM part receives the features extracted by CNN and passes the forget gate f t =σ(W f [h t-1 ,x t ]+b f ), input gate i t =σ(W i [h t-1 ,x t ]+b i ), cell state update C t =f t C t-1 +i t tanh(W C [h t-1 ,x t ]+b C ) and output gate o t =σ(W o [h t-1 ,x t ]+b o ) process data (where σ is the activation function, W f 、W i 、W C 、W o is the weight matrix, b f 、b i 、b C 、b o These parameters are trained through the back propagation algorithm and the adaptive moment estimation optimization algorithm. The accuracy of quality problem location is used as the objective function. The parameters are continuously adjusted so that the model can accurately identify the environmental factors and processing links that lead to quality problems. At the same time, the attention mechanism is introduced to calculate the attention weight of each data feature on the quality problem. (where e t =v T tanh(W a h t +b a ), v, W a 、b a , which helps the model focus more on key data and improve the accuracy of quality traceability and analysis.

[0043] The hub processing quality, processing efficiency and processing cost are taken as multi-objective functions, which are respectively expressed as f1(x), f2(x), and f3(x) as processing parameter vectors, including cutting speed, feed rate, tool selection and other parameters. A comprehensive objective function F(x)=ω1f1(x)+ω2f2(x)+ω3f3(x) is established (ω1, ω2, ω3 are the weights of each objective). In the bat algorithm, the position of the bat individual represents the processing parameter scheme. By adjusting the bat frequency f i =f min +(f max -f min )β (β is a random number), speed and location Search (where x * The search process uses an elite strategy to retain a certain number of individuals near the historical optimal solution, accelerate algorithm convergence, and iterate the algorithm to obtain the Pareto frontier solution set. Engineers select the optimal processing parameter solution from the solution set based on the company's actual needs and production conditions. For example, when the order volume is large, the processing efficiency weight is appropriately increased, and process parameters that can be produced quickly and guarantee a certain quality are given priority. When producing high-end vehicle wheels with extremely high product quality requirements, the processing quality weight is emphasized, and process parameters that can ensure high precision and high performance of the wheels are selected.

[0044] Example 2

[0045] This embodiment describes how, in a factory specializing in the production of battery trays for new energy vehicles, the entire production process has been made intelligent and efficient based on a system for improving the adaptability of the manufacturing and processing environment for new energy aluminum alloy parts.

[0046] At the processing equipment level, the stamping machine is the key equipment for the production of battery pallets. The technical staff deeply analyzes the operating conditions of the key components of the stamping machine, and accurately deploys vibration sensors and displacement sensors in locations prone to large impact and wear, such as the contact area between the punch head and the mold, the transmission connecting rod, etc., to monitor the vibration frequency and displacement change parameters of the stamping machine in real time during operation. Once abnormal fluctuations occur, the signal can be captured in time. At the component processing level, the processing tools and workpiece fixtures are crucial to the molding quality of the battery pallet. A temperature sensor is installed on the tool to monitor the temperature changes during the cutting process at all times, because excessively high temperatures will affect the tool life and processing accuracy, which may cause defects on the surface of the battery pallet. A pressure sensor is installed on the workpiece fixture to ensure that the fixture can stably fix the workpiece. To avoid displacement during the processing and affect the dimensional accuracy of the pallet, at the processing environment layer, considering that the workshop space is large and the functional areas are clearly divided, temperature and humidity sensors are deployed in the raw material storage area, because aluminum alloy raw materials are easily oxidized in a humid environment, affecting the quality of the pallet. Dust concentration sensors are deployed near the stamping area. The stamping process will produce metal debris and dust. Excessive dust concentration not only affects the health of workers, but may also affect the processing accuracy. Electromagnetic interference sensors are deployed in areas where electrical equipment is concentrated to prevent electromagnetic interference from affecting the normal operation of processing equipment and the accuracy of data collection. The deployment of sensors is not static, but adopts a hierarchical and partitioned dynamic optimization strategy, and is continuously adjusted according to the actual situation in the production process to ensure that the collected data is comprehensive, accurate and timely.

[0047] A multi-scale fusion modeling method is used to construct a digital twin model of the entire manufacturing process of the battery tray. At the microscopic scale, professional simulation software and technology are used to deeply study the characteristics of aluminum alloy materials at the microscopic level, simulate the interaction between atoms and the changes in the crystal structure during the processing, and analyze the impact of the evolution of the material microstructure on the performance of the battery tray. For example, the study studies how the changes in the atomic structure of aluminum alloys affect the strength and corrosion resistance of the tray during the high-temperature and high-pressure stamping process. At the mesoscopic scale, based on the finite element method, the local structures and key parts of the battery tray, such as reinforcement ribs and connection points, are finely modeled. By simulating different working conditions, The mechanical properties and deformation behavior of these parts can be measured, and possible quality problems such as cracks caused by stress concentration can be predicted to provide a basis for optimized design. At the macro scale, advanced 3D modeling software is used to establish a model of the entire manufacturing and processing system, including processing equipment, process flow, logistics and transportation. The entire process, from the entry of raw materials into the factory, to the transformation into finished battery trays through a series of processing steps, and then to transportation to the assembly workshop, is clearly presented in the model. Information interaction and collaboration are achieved between models of different scales through carefully designed data interfaces to ensure that the model can truly reflect the actual production process and provide reliable support for subsequent production optimization.

[0048] The large amount of data collected needs to be strictly processed before it can be used for production. The data transmission and processing module first cleans the data and adopts the abnormal data identification method based on fuzzy rough sets. The technicians divide the collected data into multiple attribute sets according to different types, such as equipment operation data, processing process data, environmental data, etc. For each attribute set, a fuzzy rough set model is constructed to identify abnormal data through specific algorithms and rules, and those data that obviously do not conform to normal production rules are eliminated to ensure the reliability of the data. Then the data is screened and the feature selection algorithm based on information entropy is used to select the most valuable information for production decision-making from the massive data features, remove redundant data, and reduce the burden of subsequent data analysis and model processing. After cleaning and screening, the data must also be normalized to unify the data into a reasonable numerical range to make it more in line with the requirements of subsequent model processing, laying the foundation for accurate analysis and decision-making.

[0049] The root cause of the quality problem of the battery tray is found by using the quality problem location model based on the improved convolutional neural network - long short-term memory network. In the CNN part, multiple convolution layers and pooling layers are designed. The convolution layer uses optimized convolution kernels. These convolution kernels are like "data filters" that can extract key local features from complex processing process data and environmental data. For example, when analyzing the welding quality problem of the battery tray, the convolution kernel can capture the characteristic information in the data such as welding temperature and welding speed. After the convolution operation extracts the features, the pooling layer is used for dimensionality reduction to reduce the amount of data while retaining important information. The LSTM part receives the features extracted by CNN and processes the data through a series of complex mechanisms such as forget gate, input gate, cell state update and output gate. The forget gate determines which Some information needs to be discarded from the cell state, the input gate controls the addition of new information, the cell state update integrates new and old information, and the output gate outputs the final result according to the updated cell state. The parameters of these gates are trained through the back propagation algorithm and the adaptive moment estimation optimization algorithm. The accuracy of quality problem positioning is used as the objective function, and the parameters are continuously adjusted to enable the model to more accurately identify the environmental factors and processing links that cause quality problems. At the same time, the attention mechanism is introduced to enable the model to pay more attention to data features that have a greater impact on quality problems when processing data, thereby improving the accuracy of quality traceability and analysis. When quality problems such as welding defects and flatness not meeting standards are found in the battery tray, the model can quickly locate the specific reasons, such as unreasonable welding process parameters or workshop humidity affecting the welding effect.

[0050] A comprehensive optimization plan is developed based on multiple objectives: processing quality, processing efficiency, and processing cost. Regarding processing quality, key indicators such as dimensional accuracy, strength, and corrosion resistance of the trays meet high standards to meet the stringent requirements of new energy vehicles for battery trays. Regarding processing efficiency, the output per unit time is increased by optimizing process flows and rationally arranging equipment operating hours to meet the growing market demand for battery trays. Regarding processing costs, raw material procurement costs, equipment energy consumption, and labor costs are strictly controlled to reduce production costs and enhance enterprise competitiveness. The decision-making and optimization module uses advanced optimization algorithms based on the results of quality traceability and analysis, combined with historical processing data and process experience, to generate a processing process optimization plan. For example, if quality traceability reveals a high number of welding defects in a batch of battery trays, and analysis shows that the cause is excessively high welding temperature, the decision-making and optimization module will refer to historical data on welding quality performance at different welding temperatures. Combined with the current equipment and raw material conditions, the module will develop an optimization plan for adjusting process parameters such as welding temperature and welding speed, and provide feedback to the processing equipment and processing environment control system. Furthermore, the module will continuously track the implementation of the optimization plan, continuously adjust and improve the plan based on actual production conditions, and ensure that the production process is always in an optimal state.

[0051] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as above in terms of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can, without departing from the scope of the technical solution of the present invention, make some changes or modifications to equivalent embodiments using the technical contents disclosed above. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A system for improving the adaptability of the manufacturing and processing environment of new energy aluminum alloy parts, characterized by: The system includes the following components: data acquisition module, digital twin model construction module, data transmission and processing module, quality traceability and analysis module, and decision-making and optimization module: The data acquisition module is used to deploy various sensors at the manufacturing site of new energy aluminum alloy parts to collect processing equipment operating parameters, parts processing process data and processing environment data; The digital twin model construction module is used to construct a digital twin model of the entire manufacturing process of new energy aluminum alloy parts based on the collected data; The data transmission and processing module transmits the collected data to the digital twin model construction module, cleans, filters and pre-processes the data, and then updates the data to the digital twin model synchronously; The quality traceability and analysis module compares the virtual quality data of parts in the digital twin model with the actual inspection quality data, and locates the environmental factors and processing links that cause quality problems in combination with the processing environment data; The decision-making and optimization module generates a processing technology optimization plan based on the quality traceability and analysis results, combined with historical processing data and process experience, and feeds back to the processing equipment and processing environment control system.

2. The system for improving the adaptability of the manufacturing and processing environment of aluminum alloy parts based on new energy according to claim 1 is characterized in that: In the data acquisition module, the deployment of sensors adopts a hierarchical and partitioned dynamic optimization strategy. At the processing equipment layer, vibration sensors and displacement sensors are deployed at the spindle and guide rail parts according to the stress and wear conditions of the key components of the equipment to collect the equipment operation vibration frequency and displacement change parameters. At the component processing process layer, temperature sensors and pressure sensors are set at the processing tools and workpiece fixtures to obtain cutting temperature and cutting force data in real time. At the processing environment layer, temperature and humidity sensors, dust concentration sensors and electromagnetic interference sensors are dispersedly deployed in different areas according to the workshop space layout and environmental characteristics. The determination of the sensor deployment location and number is optimized based on the genetic particle swarm hybrid algorithm. This algorithm combines the global search capability of the genetic algorithm and the local search advantage of the particle swarm algorithm, with minimizing the component quality fluctuation as the objective function. Where X is the sensor deployment plan vector, n is the number of quality indicators, and w i is the weight of each quality indicator, q i is the fluctuation value of the i-th quality indicator.

3. The system for improving the adaptability of aluminum alloy parts manufacturing and processing environment based on new energy according to claim 1 is characterized in that: In the digital twin model construction module, a multi-scale fusion modeling method is adopted; At the microscale, molecular dynamics simulation methods are used to construct atomic-level structural models of aluminum alloys, simulate interatomic interactions and crystal structure changes, and analyze the impact of material microstructural evolution on component performance during processing; At the mesoscopic scale, based on the finite element method, the local structure and key parts of the components are finely modeled to study their mechanical properties and deformation behavior; At the macro scale, the entire manufacturing system model is established through 3D modeling software, including processing equipment, process flow, logistics and transportation, etc. The models of each scale are connected through data interfaces to achieve information interaction and collaboration. During the model construction process, the model parameters are determined by the adaptive Kalman filter algorithm. Based on the traditional Kalman filter, the algorithm introduces the adaptive factor λ and adjusts it in real time according to the dynamic changes of the data. The state estimation equation is: in is the optimal state estimate at time k, is the predicted state at time k based on time k-1, K k is the Kalman gain, Z k is the measured value, H k is the measurement matrix.

4. The system for improving the adaptability of aluminum alloy parts manufacturing and processing environment based on new energy according to claim 1 is characterized in that: In the data transmission and processing module, data cleaning adopts the abnormal data identification method based on fuzzy rough set. First, the collected data is divided into multiple attribute sets according to different types. For each attribute set, a fuzzy rough set model is constructed and the fuzzy membership function is defined. Where a and b are parameters, and then calculate the upper approximation set of the data and lower approximation set R (X), abnormal data is satisfied The data is screened using a feature selection algorithm based on information entropy to calculate the information entropy of each data feature. where p(x i ) is the feature x i The probability of occurrence, set the information entropy threshold τ, retain the data features with information entropy greater than τ, remove redundant data, and normalize the data after cleaning and screening. Perform preprocessing.

5. The system for improving the adaptability of aluminum alloy parts manufacturing and processing environment based on new energy according to claim 1 is characterized in that: In the quality tracing and analysis module, a quality problem location model based on an improved convolutional neural network-long short-term memory network is adopted. In the CNN part, multiple convolution layers and pooling layers are designed. The convolution layer uses a custom convolution kernel K ij , its parameters are optimized by genetic algorithm to maximize the extraction of data features. For the input processing data and environmental data X, after convolution operation Extract local features and then reduce the dimension through the pooling layer. The LSTM part receives the features extracted by CNN and passes the forget gate f t =σ(W f [h t-1 , x t ]+b f ), input gate i t =σ(W i [h t-1 , x t ]+b i ), cell state update C t =f t C T-1 +i t tanh(W C [h t-1 ,x t ]+b C ) and output gate o t =σ(W o [h t-1 , x t ]+b o ) processes the data, where σ is the activation function, W f 、W i 、W C 、W o is the weight matrix, b f 、b i 、b C 、b o These parameters are trained through the back propagation algorithm and the adaptive moment estimation optimization algorithm. During the training process, the accuracy of quality problem location is used as the objective function. The parameters are adjusted so that the model can accurately identify the environmental factors and processing links that cause quality problems. At the same time, the attention mechanism is introduced to calculate the attention weight of each data feature on the quality problem. where e t =v T tanh(W a h t +b a ), v, W a 、b a As a parameter.

6. The system for improving the adaptability of aluminum alloy parts manufacturing and processing environment based on new energy according to claim 1 is characterized in that: In the decision-making and optimization module, the generation of the processing technology optimization scheme adopts an optimization method based on the multi-objective bat algorithm, with the parts processing quality, processing efficiency and processing cost as the multi-objective functions, respectively expressed as f1(X), f2(x), f3(x), where x is the processing technology parameter vector, and a comprehensive objective function F(x)=ω1f1(x)+ω2f2(x)+ω3f3(x) is established, ω1, ω2, ω3 are the weights of each objective, in the bat algorithm, the position of the individual bat represents the processing technology parameter scheme, by adjusting the bat frequency f i =f min +(f max -f min )β, β is a random number, speed and location Search for x * In order to obtain the current optimal solution, an elite strategy is introduced during the search process to retain a certain number of individuals near the historical optimal solution, thereby accelerating the convergence of the algorithm. Through algorithm iteration, the Pareto frontier solution set is obtained, and the optimal processing parameter solution is selected from the solution set based on the actual needs and production conditions of the enterprise.

7. The system for improving the adaptability of aluminum alloy parts manufacturing and processing environment based on new energy according to claim 1 is characterized in that: The system also includes an environmental intelligent control submodule, which is connected to the data acquisition module, the quality tracing and analysis module, and the decision-making and optimization module. The environmental intelligent control submodule intelligently controls the processing environment based on the environmental factors affecting the quality determined by the quality tracing and analysis module and the optimization plan generated by the decision-making and optimization module. When it is detected that the temperature is too high and affects the processing accuracy of the parts, the environmental intelligent control submodule adjusts the temperature by controlling the workshop air-conditioning system. Its control strategy adopts a dynamic temperature control algorithm based on reinforcement learning to construct a Markov decision process. The state space S is the temperature value of different areas in the workshop, and the action space A is the cooling power adjustment gear of the air-conditioning system. The reward function R is set according to the degree of influence of temperature changes on the quality of the parts. When the temperature is close to the optimal processing temperature range, a positive reward is given, and when it deviates, a negative reward is given. Through continuous interactive learning between the intelligent agent and the environment, the optimal strategy π is obtained, so that the air-conditioning system adjusts the temperature according to the optimal strategy. In the case of excessively high dust concentration, the environmental intelligent control submodule controls the operation of the dust removal equipment and adopts a fuzzy PID control algorithm to adjust the power of the dust removal equipment. According to the deviation e between the dust concentration detection value and the set value and the deviation change rate ec, the PID parameter K is output through the fuzzy controller. p , K i , K d Adjust the amount to accurately control the power of the dust removal equipment.

8. The system for improving the adaptability of aluminum alloy parts manufacturing and processing environment based on new energy according to claim 1 is characterized in that: The system also has remote collaborative operation and maintenance functions. In terms of data transmission, the high bandwidth and low latency characteristics of the 5G network are used to quickly transmit real-time data from the processing site to the remote operation and maintenance center. At the same time, blockchain technology is used to store and manage data. Each data block contains a data hash value, a timestamp, and the hash value of the previous data block. Experts at the remote operation and maintenance center can view the processing site conditions in real time through the digital twin model, and remotely evaluate and guide quality traceability and analysis results. When complex quality problems arise, experts can initiate multi-party collaborative discussions on the blockchain platform. The opinions and suggestions of each participant are recorded on the blockchain in the form of smart contracts. Based on the discussion results, experts use decision-making and optimization modules in the remote operation and maintenance center to generate optimization plans.

9. The system for improving the adaptability of aluminum alloy parts manufacturing and processing environment based on new energy according to claim 1 is characterized in that: The system also includes a knowledge graph construction and application module, which is associated with the digital twin model construction module, the quality traceability and analysis module, and the decision and optimization module. The knowledge graph construction module collects multi-source knowledge such as process knowledge, material property knowledge, equipment maintenance knowledge, etc. in the field of aluminum alloy parts manufacturing and processing, extracts, classifies and semantically analyzes the knowledge through natural language processing technology, and constructs a knowledge graph based on entities and relationships. In the quality traceability process, the quality traceability and analysis module can assist in analyzing the potential relationship between environmental factors and quality problems based on the knowledge in the knowledge graph. In the decision and optimization stage, the decision and optimization module uses the knowledge in the knowledge graph and combines the analysis results of the digital twin model to generate a more optimized processing technology plan. According to the equipment maintenance knowledge and processing technology knowledge in the knowledge graph, the equipment maintenance plan and process parameter adjustment plan are determined.

10. A method for improving the adaptability of the manufacturing and processing environment of aluminum alloy parts based on new energy, applicable to the system for improving the adaptability of the manufacturing and processing environment of aluminum alloy parts based on new energy as described in any one of claims 1 to 9, characterized in that: The specific steps of this method are: S1. Data collection step: During the manufacturing process of new energy aluminum alloy parts, the data collection module is used to collect the operating parameters of the processing equipment, the parts processing process data and the processing environment data in real time; S2. Digital twin model construction steps: Using the data acquired by the data acquisition module, a digital twin model of the entire manufacturing process of new energy aluminum alloy parts is constructed in the digital twin model construction module. The interrelationships and influencing factors between each link are fully considered during the model construction process; S3, data transmission and synchronization step: The data transmission and processing module processes the collected data and transmits it to the digital twin model construction module in real time to update various parameters in the digital twin model; S4. Quality traceability and analysis step: Periodically or when quality issues are discovered, the quality traceability and analysis module compares the virtual quality data of the components in the digital twin model with the actual quality data. When quality deviations occur, the module combines the processing environment data, applies data analysis algorithms and machine learning models, and analyzes the impact of environmental factors on quality issues. S5. Decision-making and optimization steps: The decision-making and optimization module generates a processing optimization plan based on the results of the quality traceability and analysis module, combined with historical processing data and process experience, using the optimization algorithm, and sends the plan to the processing equipment and processing environment control system to adjust the processing process.

Images