Metal material processing system

Abstract

The invention relates to the technical field of metal material processing, in particular to a metal material processing system which comprises a main processing unit, a multi-dimensional sensing unit, an autonomous moving material processing unit and a central control unit, and the central control unit comprises a signal fusion analysis module, a digital twinning module, a process dynamic optimization module and a collaborative scheduling module. The multi-dimensional sensing unit comprises a vibration sensor, an acoustic emission sensor and a high-speed vision module. According to the invention, the rejection rate is reduced, the consistency of the quality of batch machined parts is ensured, seamless cooperation of machining and logistics is realized, the equipment waiting time is reduced to the greatest extent, the overall equipment utilization rate is improved, the data processing pressure of the central control unit is effectively reduced, and the response speed of the system to real-time events can be improved; and the real-time performance and the stability of the system are enhanced.

Description

Technical Field

[0001] This invention relates to the field of metal material processing technology, specifically to a metal material processing system. Background Technology

[0002] Metal processing, as a core link in modern manufacturing, encompasses a wide range of precision and ultra-precision machining processes, including turning, milling, and additive manufacturing. With the increasing demands for performance and reliability of complex components in high-end equipment, aerospace, and precision instruments, unprecedented high requirements are being placed on the precision, efficiency, consistency, and adaptability of metal processing.

[0003] Currently, existing metal processing systems typically rely on pre-set fixed programs (such as CNC code) to control machine tools to perform machining operations, and then use offline inspection methods (such as coordinate measuring machines) to perform random or full inspections of workpiece quality after machining. This approach has significant drawbacks: First, the physical states during machining (such as tool wear, sudden changes in cutting force, thermal deformation, vibration, and chatter) cannot be fully and realistically perceived and understood, resulting in a narrow process window. This makes it difficult to maximize machining efficiency while ensuring quality, and machining anomalies and defects are often only discovered after they occur, leading to wasted materials and time. Second, although some research has attempted to introduce single sensors (such as vibration sensors or vision sensors) for online monitoring, metal processing is a complex process involving strong coupling of multiple physical fields (force, heat, sound, and light). A single signal source cannot comprehensively and accurately characterize the entire machining state, let alone achieve precise location and early warning of defect roots.

[0004] Furthermore, existing metal processing systems suffer from limited intelligence and insufficient coordination among subsystems. Material handling (loading, unloading, and work-in-process transfer) typically relies on manual labor or simple automated equipment (such as fixed robotic arms or conveyor belts), lacking real-time information interaction and dynamic scheduling with core processing procedures. This hinders flexible responses to changes in processing plans or abnormal situations, limiting the overall flexibility and efficiency of the production line. Simultaneously, with the rise of digital twin technology, its application in the processing field is largely limited to offline process simulation and optimization, or simple equipment status visualization, failing to achieve deep closed-loop integration with real-time processing data. This results in a disconnect between the simulation model and the physical world, preventing it from realizing its enormous potential in real-time decision support. No solutions have yet been proposed to address these technical issues. Summary of the Invention

[0005] In response to the problems in related technologies, this invention proposes a metal material processing system to overcome the aforementioned technical problems existing in the prior art. The purpose of this invention is to reduce the scrap rate, ensure the consistency of batch processed parts quality, achieve seamless coordination between processing and logistics, minimize equipment waiting time, improve overall equipment utilization, effectively reduce the data processing pressure on the central control unit, and at the same time improve the system's response speed to real-time events, thereby enhancing the system's real-time performance and stability.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a metal material processing system, comprising: The main processing unit includes processing equipment for performing at least one metal processing operation; A multi-dimensional sensing unit, integrated on the main processing unit, is used to collect multi-physical field signals during the processing in real time. The multi-physical field signals include at least vibration signals, acoustic emission signals, and optical image signals. An autonomous mobile material handling unit includes a mobile platform, a material clamping mechanism disposed on the mobile platform, and a first navigation module, for autonomous transfer and loading / unloading within the processing area; The central control unit is communicatively connected to the main processing unit, the multi-dimensional sensing unit, and the autonomous mobile material handling unit, respectively. The central control unit includes: The signal fusion analysis module is used to synchronously fuse the multi-physics field signals to generate a comprehensive quality index characterizing the processing status. The digital twin module has a built-in parametric virtual model of the processing equipment and the workpiece. The parametric virtual model can perform synchronous simulation based on the real-time control parameters of the main processing unit and the comprehensive quality index. The process dynamic optimization module is used to compare the comprehensive quality index with a preset threshold, and based on the simulation results of the digital twin module, generate correction instructions for the control parameters of the main processing unit or adjustment strategies for subsequent processing paths in real time. The collaborative scheduling module is used to issue scheduling commands, including target location and task instructions, to the autonomous moving material handling unit based on the processing status and task queue of the main processing unit.

[0007] Preferably, the multidimensional sensing unit includes: Vibration sensors are used to collect structural vibration signals in the processing area; Acoustic emission sensors are used to collect stress wave signals generated by microscopic deformation of materials; A high-speed vision module is used to acquire dynamic images of the processing area; The signal fusion analysis module employs a timestamp-aligned data fusion algorithm to correlate and analyze the vibration signal, acoustic emission signal, and optical image signal in the time and spatial domains to identify and locate processing defects.

[0008] Preferably, the signal fusion analysis module extracts the temperature field distribution of the processing area from the image acquired by the high-speed vision module, and combines the spectral characteristics of the vibration signal and the amplitude and counting characteristics of the acoustic emission signal to establish a multivariate regression model associated with material removal rate, tool wear state and workpiece surface integrity. The multivariate regression model is used to calculate the comprehensive quality index based on the real-time fused multiphysics field signal. The comprehensive quality index includes at least the predicted surface roughness value, residual stress level and tool health index.

[0009] Preferably, the digital twin module includes: The physical property layer includes the material constitutive model and the tool geometry and wear model. The dynamic simulation layer can simulate cutting forces and thermal deformation processes; A real-time data mapping interface is used to input the real-time control parameters of the main processing unit and the comprehensive quality index as boundary conditions into the dynamic simulation layer; When the comprehensive quality index deviates from the preset threshold, the process dynamic optimization module triggers the digital twin module to conduct multiple sets of simulation experiments to obtain a set of control parameter corrections that allow the comprehensive quality index to return to the target range.

[0010] Preferably, when the comprehensive quality index indicates that the tool wear is aggravated, the process dynamic optimization module generates instructions to adjust the spindle speed, feed rate and enable the backup tool path; when the comprehensive quality index indicates that the local heat accumulation of the workpiece exceeds the limit, the process dynamic optimization module generates instructions to adjust the coolant injection parameters, insert a cooling dwell period and adjust the machining sequence.

[0011] Preferably, the autonomous mobile material handling unit further includes: The second navigation module is used to achieve precise positioning and obstacle avoidance of the mobile platform within the processing area; A force sensing module, integrated into the material clamping mechanism, is used to detect clamping force and placement force; The task instructions issued by the collaborative scheduling module include force-controlled feeding, force-controlled unloading, and in-machine detection auxiliary tasks based on feedback from the force sensing module.

[0012] Preferably, it also includes an edge computing gateway, which is located on one side of the main processing unit, for preprocessing and feature extraction of the raw high-bandwidth data collected by the multi-dimensional sensing unit, and then uploading the dimensionality-reduced feature data to the central control unit.

[0013] Preferably, it also includes a global resource management platform, which is communicatively connected to the central control unit. The global resource management platform stores raw material inventory information, tool library information, historical processing technology packages, and corresponding comprehensive quality index files. When the process dynamic optimization module generates correction instructions, it calls the historical data in the global resource management platform for analog learning and verification.

[0014] Preferably, the main machining unit is one of a CNC milling center, a turning-milling composite machining center, or a laser additive manufacturing equipment.

[0015] Compared with the prior art, the beneficial effects of the present invention are: (1) This invention is a metal material processing system. Through the multi-physics field signal fusion analysis of the multi-dimensional sensing unit, it can penetrate the limitations of a single signal and associate different physical phenomena in the time and spatial domains. It can not only detect early defects such as tool micro-breakage and workpiece surface micro-cracks earlier, but also achieve precise positioning of defect sources, providing precise targets for subsequent process adjustments. The multivariate regression model established by the signal fusion analysis module can predict key quality indicators and tool health status based on real-time data, reduce scrap rate, and ensure high consistency of batch processed parts quality. (2) The present invention is a metal material processing system. The digital twin module can receive real-time data and synchronously simulate the twin. When the quality index is detected to deviate, the process dynamic optimization module can conduct simulation experiments in virtual space to find the optimal process parameter correction set. It can actively deal with disturbances such as tool wear and material inhomogeneity, and always keep the processing process in the optimal state. Combined with the historical process and quality archives in the global resource management platform, the system can perform analog learning. (3) This invention is a metal material processing system. The autonomous mobile material handling unit is integrated into the processing control closed loop through the collaborative scheduling module. The material flow is dynamically scheduled according to the real-time status of the main processing unit and the task queue, realizing seamless collaboration between processing and logistics, minimizing equipment waiting time and improving overall equipment utilization. The force sensing module of the autonomous mobile material handling unit effectively protects precision workpieces and machine tool fixtures, and expands on-machine detection, enhancing the automation and intelligence level of the system. (4) This invention is a metal material processing system. By setting up an edge computing gateway, the high-bandwidth raw data generated by the sensor is preprocessed and features are extracted locally, which effectively reduces the data processing pressure of the central control unit, reduces the network bandwidth requirement, and improves the system's response speed to real-time events, thereby enhancing the system's real-time performance and stability. The main processing unit can be adapted to various mainstream metal processing equipment such as CNC milling, mill-turning composite, and additive manufacturing. The multi-dimensional sensing unit can be configured and expanded according to specific needs, and has good universality and scalability, making it easy to deploy and apply in manufacturing scenarios of different scales and types. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall framework structure of the present invention. Detailed Implementation

[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0018] Example Please see Figure 1 This invention proposes a technical solution for a metal material processing system: a metal material processing system comprising: The main machining unit includes machining equipment for performing at least one metal processing operation; specifically, the main machining unit is the core of the system and is responsible for performing various metal processing operations, such as milling, turning and laser additive manufacturing. The main machining unit is equipped with advanced CNC technology to ensure machining accuracy and repeatability. At the same time, the main machining unit integrates multiple machining equipment, enabling it to flexibly handle different types of machining tasks. The multi-dimensional sensing unit, integrated into the main machining unit, is used to acquire multi-physics field signals in real time during the machining process. These signals include at least vibration signals, acoustic emission signals, and optical image signals. Specifically, vibration signals are acquired by vibration sensors, reflecting the equipment's operating status and the tool's working condition. Acoustic emission signals are collected by acoustic emission sensors, indicating stress changes within the material and helping to identify microscopic defects. Optical image signals are acquired by a high-speed vision module, providing real-time images of the machined surface for easy surface quality inspection. The signal fusion analysis module uses a timestamp-aligned data fusion algorithm to synchronously analyze vibration signals, acoustic emission signals, and optical image signals to locate and identify machining defects, ensuring the stability and reliability of the machining process. The autonomous mobile material handling unit includes a mobile platform, a material clamping mechanism mounted on the mobile platform, and a first navigation module, used for autonomous transfer and loading / unloading within the processing area. Specifically, the autonomous mobile material handling unit is mainly responsible for autonomous transfer and loading / unloading within the processing area. The mobile platform can move flexibly within the processing area according to task requirements. The material clamping mechanism is used to safely and effectively clamp and place workpieces, and integrates a force sensing module to monitor the clamping and placing force to ensure operational safety. The first navigation module achieves precise positioning and obstacle avoidance, ensuring efficient material transfer. The central control unit is communicatively connected to the main processing unit, the multi-dimensional sensing unit, and the autonomous mobile material handling unit, respectively. The central control unit includes: The signal fusion analysis module is used to synchronously fuse multi-physics field signals to generate a comprehensive quality index characterizing the processing status. The digital twin module contains a parametric virtual model of the processing equipment and the workpiece. The parametric virtual model can perform synchronous simulation based on the real-time control parameters and comprehensive quality indicators of the main processing unit. The process dynamic optimization module is used to compare the comprehensive quality index with the preset threshold and, based on the simulation results of the digital twin module, generate correction instructions for the control parameters of the main processing unit or adjustment strategies for subsequent processing paths in real time. The collaborative scheduling module issues scheduling commands, including target locations and task instructions, to the autonomous material handling unit based on the processing status and task queue of the main processing unit. Specifically, the signal fusion analysis module integrates multi-physics field signals from different sources to generate a comprehensive quality index characterizing the processing status for subsequent dynamic optimization and adjustment. The digital twin module establishes a virtual model of the processing equipment and workpiece, and performs simulation and prediction by comparing real-time data with the model to support the decision-making process. The process dynamic optimization module corrects processing control parameters and adjusts the processing path in real time, and optimizes the processing process based on the comparison results of the comprehensive quality index and preset thresholds. The collaborative scheduling module generates optimal scheduling instructions for the autonomous material handling unit based on the current processing status and task queue to improve overall processing efficiency. The signal fusion analysis module extracts the temperature field distribution from the vision module, combines it with vibration and acoustic emission signals to establish a multivariate regression model, and correlates material removal rate, tool wear status, and workpiece surface integrity. The comprehensive quality index includes not only the predicted surface roughness value, but also the residual stress level and tool health index, helping operators to monitor and adjust in real time during the processing.

[0019] Furthermore, the multidimensional sensing unit includes: Vibration sensors are used to collect structural vibration signals in the processing area; Acoustic emission sensors are used to collect stress wave signals generated by microscopic deformation of materials; A high-speed vision module is used to acquire dynamic images of the processing area; The signal fusion analysis module employs a timestamp-aligned data fusion algorithm to correlate vibration signals, acoustic emission signals, and optical image signals in the time and spatial domains to identify and locate processing defects.

[0020] Furthermore, the signal fusion analysis module extracts the temperature field distribution of the processing area from the images acquired by the high-speed vision module. Combining the spectral characteristics of the vibration signal and the amplitude and counting characteristics of the acoustic emission signal, it establishes a multivariate regression model that is associated with the material removal rate, tool wear status and workpiece surface integrity. The multivariate regression model is used to calculate the comprehensive quality index based on the real-time fused multi-physics field signals. The comprehensive quality index includes at least the predicted surface roughness value, residual stress level and tool health index.

[0021] Furthermore, the digital twin module includes: The physical property layer includes the material constitutive model and the tool geometry and wear model. The dynamic simulation layer can simulate cutting forces and thermal deformation processes; The real-time data mapping interface is used to input the real-time control parameters and comprehensive quality indicators of the main processing unit as boundary conditions into the dynamic simulation layer; Among them, when the comprehensive quality index deviates from the preset threshold, the process dynamic optimization module triggers the digital twin module to conduct multiple sets of simulation experiments to obtain a set of control parameter corrections that allow the comprehensive quality index to return to the target range.

[0022] Furthermore, when the overall quality index indicates that tool wear is intensifying, the process dynamic optimization module generates instructions to adjust the spindle speed, feed rate, and enable the backup toolpath. When the overall quality index indicates that local heat accumulation in the workpiece exceeds the limit, the process dynamic optimization module generates instructions to adjust the coolant injection parameters, insert a cooling dwell period, and adjust the machining sequence.

[0023] In this embodiment, when the comprehensive quality index deviates from the preset threshold, the process dynamic optimization module will trigger the digital twin module to conduct simulation experiments to find a suitable set of control parameter corrections to ensure the stability of the processing and product quality. For example, in the case of accelerated tool wear or excessive local heat accumulation on the workpiece, the system will automatically generate corresponding adjustment instructions to ensure the continuity and safety of processing.

[0024] Furthermore, the autonomous mobile material handling unit also includes: The second navigation module is used to achieve precise positioning and obstacle avoidance of the mobile station within the processing area; The force sensing module, integrated into the material clamping mechanism, is used to detect clamping force and placement force; Among them, the task instructions issued by the collaborative scheduling module include force-controlled feeding, force-controlled unloading, and in-machine detection auxiliary tasks based on feedback from the force sensing module.

[0025] Furthermore, it also includes an edge computing gateway, which is set on one side of the main processing unit to preprocess and extract features from the raw high-bandwidth data collected by the multi-dimensional sensing unit, and then upload the dimensionality-reduced feature data to the central control unit.

[0026] In this embodiment, the edge computing gateway is located next to the main processing unit and is responsible for preprocessing and feature extraction of the high-bandwidth raw data from the multi-dimensional sensing unit to reduce the data transmission burden and improve real-time response capabilities. This processing step ensures that important data can quickly reach the central control unit.

[0027] Furthermore, it also includes a global resource management platform that communicates with the central control unit. The global resource management platform stores raw material inventory information, tool library information, historical processing technology packages, and corresponding comprehensive quality index files. When generating correction instructions, the process dynamic optimization module calls on historical data in the global resource management platform for analog learning and verification.

[0028] In this embodiment, the global resource management platform stores rich resource information, including raw material inventory, tool information, and historical processing technology packages. The process dynamic optimization module can use this historical data for analog learning to optimize the current processing strategy and ensure efficient use of resources.

[0029] Furthermore, the main machining unit is one of the following: CNC milling center, mill-turn machining center, or laser additive manufacturing equipment.

[0030] In this embodiment, the spindle speed range of the CNC milling center is 50-30000 rpm, supporting constant surface speed cutting, positioning accuracy ≤ ±0.003 mm, repeatability ≤ ±0.001 mm, supporting five-axis linkage machining capability, and equipped with a high-pressure cooling system (pressure ≥ 70 bar); the spindle and sub-spindle synchronization accuracy of the mill-turn machining center is ≤ ±0.002 mm, B-axis rotation resolution ≤ 0.001°, tool magazine capacity ≥ 40 tools, supporting random tool changing; equipped with an automatic tool setter and tool breakage detection device; the laser power stability of the laser additive manufacturing equipment is ≤ ±1%, powder thickness control accuracy is ≤ ±0.01 mm, oxygen content control in the forming chamber is ≤ 50 ppm, and the substrate preheating temperature is up to 600℃ with an accuracy of ±5℃. The CNC milling center, mill-turn machining center, and laser additive manufacturing equipment support protocols such as OPC UA and MTConnect, support G-code extended instructions and macro program calls, and provide real-time feedback on parameters such as spindle load, axis position, and temperature.

[0031] Working principle of the invention: The central control unit's collaborative scheduling module receives processing task instructions from the production management system and initiates the entire system's workflow: it breaks down processing orders into specific process sequences, technological requirements, and quality indicators; it queries the global resource management platform to confirm the availability of required tools, fixtures, raw materials, and other resources; it plans a safe and efficient transportation path from the material area to the processing area for the autonomous mobile material handling unit; it calls up historically similar process packages as initial processing parameters; it sends these parameters to the main processing unit and the digital twin module; it initiates the preheating program of the main processing unit; and it activates the self-checking function of the multi-dimensional sensing unit.

[0032] During the machining process, the multi-dimensional sensing unit synchronously acquires multi-source data at high frequency: vibration sensors (20kHz), acoustic emission sensors (1MHz bandwidth), and high-speed vision modules (1000fps) achieve μs-level time synchronization based on the IEEE 1588 precision time protocol, ensuring data spatiotemporal alignment. Vibration signals monitor the mechanical vibration of the spindle, tool, and worktable, identifying anomalies such as chatter and imbalance. Acoustic emission signals capture stress waves from microscopic material deformation and crack initiation, enabling early warning of defects. Vision signals acquire optical and thermal images of the machining area, analyzing chip morphology, temperature distribution, and surface texture. The edge computing gateway processes the raw high-bandwidth data in real time: implementing digital filtering (bandpass, notch filtering) to eliminate noise interference, extracting time-domain features (root mean square value, peak factor, kurtosis), performing frequency-domain transformation (FFT, wavelet packet decomposition) to obtain spectral features, extracting thermal features such as temperature gradient and heat accumulation rate from thermal imaging, and compressing high-dimensional feature vectors to 32-dimensional core features through principal component analysis, reducing data transmission volume to 10% of the original.

[0033] The signal fusion analysis module employs a three-level fusion architecture for in-depth analysis: Based on timestamps, it registers spatiotemporally aligned data from vibration, acoustic emission, and vision to form a unified spatiotemporal reference system. Canonical correlation analysis is used to uncover deep correlations across modal features, establishing a multivariate regression model between vibration spectrum, acoustic emission amplitude / count, and visual temperature field. This model maps: {vibration energy, AE event rate, temperature gradient} → {predicted surface roughness, residual stress level, tool health index}. DS evidence theory and Bayesian networks are applied to synthesize independent judgments from different sensors, calculating the overall confidence level of the machining state and outputting a comprehensive quality index (QPI). The comprehensive quality index includes: Quantitative quality parameters: surface roughness Ra (μm), residual stress level (level 1-5); Process health: Tool health index (0-100%), machining stability index (0-1); Defect risks: flutter risk factor, thermal damage risk factor.

[0034] The digital twin module constructs and updates virtual machining scenarios based on real-time data: Through a real-time data mapping interface, the real-time control parameters (spindle speed, feed rate, depth of cut) and comprehensive quality indicators of the main machining unit are input into the virtual model as boundary conditions. The dynamic simulation layer performs cutting force calculation (based on the improved Oxley model) and thermo-mechanical coupled finite element analysis to predict workpiece thermal deformation and stress distribution. The wear evolution simulation updates the tool geometry and wear model according to the machining history and current working conditions to predict the remaining tool life. It simulates the machining process 5-30 seconds ahead in the virtual space to predict the quality deviations and risks that may occur under the current parameters. The simulation prediction results are compared with the actual values ​​of the comprehensive quality indicators to calibrate model errors and ensure the consistency between the digital twin and the physical world.

[0035] The process dynamic optimization module makes intelligent decisions based on perception and simulation results: Status Judgment: Compare the real-time value of the comprehensive quality indicator with a preset threshold range to identify the following situations: Normal state: All indicators are in the green zone, and the system maintains the current parameters; Warning status: When a single indicator enters the yellow zone, Level 1 optimization is initiated; Abnormal status: Multiple indicators enter the red zone or a single indicator exceeds the limit severely, and emergency optimization is initiated.

[0036] Optimized triggering: When an abnormal or warning state is detected, the digital twin module is triggered to perform a virtual test field simulation. Under the premise of ensuring the constraints (machine tool power, tool strength, dimensional tolerance); Multiple sets of candidate parameter combinations are generated quickly using a multi-objective genetic algorithm (NSGA-II). Simulate the processing results under various parameter sets in parallel within a virtual environment; The candidate solutions are evaluated based on the objective function J = w1·f(quality deviation) + w2·f(processing time) + w3·f(energy consumption), and the Pareto optimal solution is selected.

[0037] Instruction generation: Generates specific, executable adjustment instructions, including: Parameter fine-tuning: incremental adjustments such as spindle speed ±5% and feed rate ±10%; Path replanning: Enable alternative toolpaths and adjust toolpath strategies; Auxiliary control: Adjust coolant injection parameters (flow rate, pressure, direction) and insert cooling residence cycle; Resource Request: When the tool health index is below the threshold, request the autonomous moving unit to replace the tool.

[0038] The collaborative scheduling module enables deep collaboration between processing operations and material flow: Instruction distribution: Simultaneously issue process optimization instructions to: Main machining unit: Adjust CNC program parameters; Autonomous mobile material handling unit: Prepares materials needed for the next process or performs auxiliary tasks.

[0039] Material scheduling: Predicting material demand based on real-time processing status. Proactive scheduling: 3-5 minutes before processing is completed, the scheduling material unit will transport the next workpiece to the standby area; Abnormal response scheduling: When a tool needs to be replaced, the material unit is immediately scheduled to retrieve a tool from the tool magazine; Force-controlled precision operation: The force sensing module of the material unit ensures safe and precise operation. Adaptive clamping: Automatically adjusts the clamping force (range 5-500N) according to the workpiece weight and material. Compliant placement: Admittance control is used when placing workpieces to achieve "zero impact" docking; In-machine inspection assistance: Transport the measuring probe to the designated position to assist in completing automated measurement; Dynamic obstacle avoidance and replanning: The second navigation module monitors environmental changes in real time and immediately replans the route when encountering sudden obstacles to ensure the continuity of transportation.

[0040] The system continuously verifies the adjustment effects and accumulates process knowledge: Effect Verification: After process adjustments, the multi-dimensional sensing unit immediately monitors the changing trends of quality indicators; verifying the effectiveness of the adjustments: If the indicator returns to the green zone, record this optimization as a successful case. If the indicators do not improve or worsen, initiate a second round of optimization or upgrade to manual intervention; Process logging: The edge computing gateway and central control unit record synchronously: Complete raw signal data and extracted features; Each optimization includes its antecedents (abnormal states), decision-making process (candidate solution evaluation), and consequences (adjustment effects). Equipment status parameters and resource consumption data; Knowledge Updates: The global resource management platform continuously learns based on new data. Incremental model training: updating the multivariate regression model with new data to improve prediction accuracy; Case library expansion: Successful optimization cases will be stored in the knowledge base as a five-tuple of {material-tool-feature-problem-solution}; Rule mining: By analyzing association rules, implicit relationships between process parameters are discovered, and new expert rules are formed; Adaptive evolution: The system can identify recurring abnormal patterns, automatically optimize preset thresholds and algorithm parameters, and achieve continuous self-improvement.

[0041] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A metal material processing system, characterized in that, include: The main processing unit includes processing equipment for performing at least one metal processing operation; A multi-dimensional sensing unit, integrated on the main processing unit, is used to collect multi-physical field signals during the processing in real time. The multi-physical field signals include at least vibration signals, acoustic emission signals, and optical image signals. An autonomous mobile material handling unit includes a mobile platform, a material clamping mechanism disposed on the mobile platform, and a first navigation module, for autonomous transfer and loading / unloading within the processing area; The central control unit is communicatively connected to the main processing unit, the multi-dimensional sensing unit, and the autonomous mobile material handling unit, respectively. The central control unit includes: The signal fusion analysis module is used to synchronously fuse the multi-physics field signals to generate a comprehensive quality index characterizing the processing status. The digital twin module has a built-in parametric virtual model of the processing equipment and the workpiece. The parametric virtual model can perform synchronous simulation based on the real-time control parameters of the main processing unit and the comprehensive quality index. The process dynamic optimization module is used to compare the comprehensive quality index with a preset threshold, and based on the simulation results of the digital twin module, generate correction instructions for the control parameters of the main processing unit or adjustment strategies for subsequent processing paths in real time. The collaborative scheduling module is used to issue scheduling commands, including target location and task instructions, to the autonomous moving material handling unit based on the processing status and task queue of the main processing unit.

2. The metal material processing system according to claim 1, characterized in that: The multidimensional sensing unit includes: Vibration sensors are used to collect structural vibration signals in the processing area; Acoustic emission sensors are used to collect stress wave signals generated by microscopic deformation of materials; A high-speed vision module is used to acquire dynamic images of the processing area; The signal fusion analysis module employs a timestamp-aligned data fusion algorithm to correlate and analyze the vibration signal, acoustic emission signal, and optical image signal in the time and spatial domains to identify and locate processing defects.

3. The metal material processing system according to claim 2, characterized in that: The signal fusion analysis module extracts the temperature field distribution of the processing area from the image acquired by the high-speed vision module. Combining the spectral characteristics of the vibration signal and the amplitude and counting characteristics of the acoustic emission signal, it establishes a multivariate regression model associated with material removal rate, tool wear state and workpiece surface integrity. The multivariate regression model is used to calculate the comprehensive quality index based on the real-time fused multiphysics field signal. The comprehensive quality index includes at least the predicted surface roughness value, residual stress level and tool health index.

4. The metal material processing system according to claim 1, characterized in that: The digital twin module includes: The physical property layer includes the material constitutive model and the tool geometry and wear model. The dynamic simulation layer can simulate cutting forces and thermal deformation processes; A real-time data mapping interface is used to input the real-time control parameters of the main processing unit and the comprehensive quality index as boundary conditions into the dynamic simulation layer; When the comprehensive quality index deviates from the preset threshold, the process dynamic optimization module triggers the digital twin module to conduct multiple sets of simulation experiments to obtain a set of control parameter corrections that allow the comprehensive quality index to return to the target range.

5. A metal material processing system according to claim 4, characterized in that: When the comprehensive quality index indicates that tool wear is increasing, the process dynamic optimization module generates instructions to adjust the spindle speed, feed rate, and enable the backup toolpath. When the comprehensive quality index indicates that the local heat accumulation of the workpiece exceeds the limit, the process dynamic optimization module generates instructions to adjust the coolant injection parameters, insert a cooling dwell period, and adjust the machining sequence.

6. The metal material processing system according to claim 1, characterized in that: The autonomous mobile material handling unit also includes: The second navigation module is used to achieve precise positioning and obstacle avoidance of the mobile platform within the processing area; A force sensing module, integrated into the material clamping mechanism, is used to detect clamping force and placement force; The task instructions issued by the collaborative scheduling module include force-controlled feeding, force-controlled unloading, and in-machine detection auxiliary tasks based on feedback from the force sensing module.

7. A metal material processing system according to claim 1, characterized in that: It also includes an edge computing gateway, which is located on one side of the main processing unit, for preprocessing and feature extraction of the raw high-bandwidth data collected by the multi-dimensional sensing unit, and then uploading the dimensionality-reduced feature data to the central control unit.

8. A metal material processing system according to claim 1, characterized in that: It also includes a global resource management platform, which is connected to the central control unit. The global resource management platform stores raw material inventory information, tool library information, historical processing technology packages and corresponding comprehensive quality index files. When the process dynamic optimization module generates correction instructions, it calls the historical data in the global resource management platform for analog learning and verification.

9. A metal material processing system according to claim 1, characterized in that: The main machining unit is one of the following: CNC milling center, mill-turn machining center, or laser additive manufacturing equipment.

Images