Distributed manufacturing resource intelligent scheduling method and system

Abstract

The invention relates to the technical field of distributed manufacturing resource intelligent scheduling, in particular to a distributed manufacturing resource intelligent scheduling method and system, and the method comprises the following steps: obtaining equipment operation state data and workpiece quality data; based on the equipment operation state data and the workpiece quality data, evaluating the equipment performance degradation degree; correcting equipment capability parameters in the scheduling system according to the equipment performance degradation degree; before task allocation, task risk assessment information is generated based on the equipment performance degradation degree and task requirements; and according to the task risk assessment information, assisting a dispatcher to make a task allocation decision. The intelligent level of the scheduling system and the accuracy of decision making can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent scheduling of distributed manufacturing resources, and particularly relates to an intelligent scheduling method and system for distributed manufacturing resources. BACKGROUND

[0002] In modern industrial production systems, managing geographically dispersed and diverse production facilities to improve efficiency, control costs and ensure on-time delivery is a major challenge. Intelligent scheduling methods for distributed manufacturing resources are widely used to solve these problems, and the core is to efficiently and reasonably allocate tasks and arrange production processes. However, in many actual deployed distributed manufacturing systems, the running state information of the core processing equipment as a key input of the intelligent scheduling method often has inherent limitations in its updating mechanism. These state updates mainly rely on hard fault alarm triggers built into the equipment, or periodic manual inspection and data entry by on-site operators. This dependence causes the scheduling system's awareness of the device state to be potentially lagging, making it difficult to reflect subtle changes in device operation in real time, especially when device performance is gradually deteriorating. SUMMARY

[0003] The present application aims to address the above-mentioned deficiencies by providing an intelligent scheduling method and system for distributed manufacturing resources.

[0004] The present application adopts the following technical solutions: An intelligent scheduling method for distributed manufacturing resources, the method comprising the following steps: Obtaining device running state data and workpiece quality data; Based on the device running state data and the workpiece quality data, evaluating the degree of device performance degradation; According to the degree of device performance degradation, correcting the device capability parameters in the scheduling system; Before task allocation, generating task risk assessment information based on the degree of device performance degradation and task requirements; According to the task risk assessment information, assisting the scheduler in making task allocation decisions; The step of evaluating the degree of device performance degradation based on the device running state data and the workpiece quality data comprises: obtaining multi-source machining process signals of the device running, processing the multi-source machining process signals of the device running to obtain real-time machining response features, obtaining material response feature patterns of new alloy materials, the material response feature patterns of new alloy materials representing the interaction between device implicit degradation and material characteristics leading to subsurface microscopic defects, comparing the real-time machining response features with the material response feature patterns of new alloy materials to predict internal defect risks, and comprehensively evaluating the degree of device performance degradation by combining the internal defect risks with the device vibration deviation degree and the workpiece surface quality deviation degree.

[0005] By the technical solution, the present application can perceive the implicit degradation of the device performance in real time and incorporate it into the scheduling decision, thereby avoiding scheduling errors caused by lagging device state perception and effectively solving the problems of untimely device state updating, lack of correlation analysis between quality data and device state data in the prior art, and significantly improving the intelligent level of the scheduling system and the accuracy of the decision.

[0006] Further, the step of correcting the device capability parameter in the scheduling system according to the device performance degradation degree comprises: obtaining microstructure feature data and mechanical property data of the current batch of new materials; comparing the microstructure feature data and the mechanical property data of the current batch of new materials with the preset material property benchmark to identify the property deviation degree of the current batch of materials; dynamically adjusting a weight coefficient or an adjustment factor for correcting the device capability parameter according to the property deviation degree of the current batch of materials; correcting the device capability parameter in the scheduling system by using the adjusted weight coefficient or adjustment factor in combination with the device performance degradation degree.

[0007] Further, before task allocation, the step of generating task risk assessment information based on the device performance degradation degree and the task requirements comprises: identifying the position of the task to be allocated in the product production process and the dependency relationship with subsequent processes; predicting the delay time or the quality unqualified rate that may be generated when the task to be allocated is executed on the device; calculating the influence of the delay time or the quality unqualified rate on the start time, the completion time of the subsequent processes and the product delivery cycle according to the position of the task to be allocated in the product production process and the dependency relationship with subsequent processes; quantifying the influence as a production chain transmission effect risk score and integrating it into the task risk assessment information.

[0008] Further, the step of obtaining the device running state data and the workpiece quality data comprises: deploying high-sensitivity piezoelectric acceleration sensors and high-precision acoustic emission sensors at key positions of the core machining device to collect micro-vibration signals and acoustic emission signals generated during the operation of the device; performing real-time digital filtering processing on the micro-vibration signals and the acoustic emission signals; performing time domain feature extraction and frequency domain feature extraction on the filtered micro-vibration signals; performing time domain feature extraction and frequency domain feature extraction on the filtered acoustic emission signals; The extracted vibration features and acoustic emission features are fused to form the equipment operation state data; An online optical detection system and a three-coordinate measuring machine are deployed at the end of the production line to detect each machined workpiece in real time and obtain size deviation and surface roughness data of the workpiece, wherein the size deviation and surface roughness data of the workpiece are workpiece quality data; The real-time detected workpiece quality data is accurately associated with the equipment operation state data for machining the workpiece through the unique identification of the workpiece.

[0009] Further, the step of fusing the extracted vibration features and acoustic emission features to form the equipment operation state data comprises: obtaining current working mode information of the equipment; According to the current working mode information of the equipment, the vibration feature fusion weight and the acoustic emission feature fusion weight corresponding to the current working mode are obtained from the preset working mode and fusion weight mapping rule; The vibration features and acoustic emission features are fused by using the vibration feature fusion weight and the acoustic emission feature fusion weight to form the equipment operation state data.

[0010] Further, the step of fusing the extracted vibration features and acoustic emission features to form the equipment operation state data comprises: When the working mode of the equipment is switched, the working mode information before switching and the working mode information after switching are obtained; According to the working mode information before switching and the working mode information after switching, the transition phase of the working mode switching is determined; In the transition phase of the working mode switching, a time decay-based weighted fusion strategy is used to interpolate the fusion weight corresponding to the working mode before switching and the fusion weight corresponding to the working mode after switching to obtain a transition fusion weight; The vibration features and acoustic emission features are fused by using the transition fusion weight to form the equipment operation state data.

[0011] Further, before task allocation, the step of generating task risk assessment information based on the equipment performance degradation degree and the task requirements comprises: obtaining a specific index threshold of the product microstructure or surface topography in the task requirements; According to the equipment performance degradation degree and the specific index threshold, the risk level of subsurface defect types and internal stress concentration generated in the machining process is predicted; Obtaining the association rule between the long-term reliability or specific function of the product and the risk level of subsurface defect types and internal stress concentration; According to the association rule, the influence of the sub-surface defect type and the internal stress concentration risk level on the long-term reliability or specific function of the product is quantified; The quantified influence is taken as the hidden defect risk score and integrated into the task risk assessment information.

[0012] Further, the step of integrating the quantified influence as the hidden defect risk score into the task risk assessment information comprises: obtaining hidden defect type information contained in the quantified influence; According to the hidden defect type information, a risk presentation granularity or dimension corresponding to the hidden defect type is obtained from a preset defect type and risk presentation rule; According to the risk presentation granularity or dimension, the quantified influence is converted to generate a hidden defect risk score with a specific granularity or dimension; The hidden defect risk score with the specific granularity or dimension is integrated into the task risk assessment information.

[0013] The application also discloses a distributed manufacturing resource intelligent scheduling system applied to the distributed manufacturing resource intelligent scheduling method, and the system comprises: An acquisition module is configured to acquire equipment operation state data and workpiece quality data; A performance evaluation module is configured to evaluate the equipment performance degradation degree based on the equipment operation state data and the workpiece quality data; A correction module is configured to correct the equipment capability parameters in the scheduling system according to the equipment performance degradation degree; A risk assessment module is configured to generate task risk assessment information based on the equipment performance degradation degree and task requirements before task allocation; An auxiliary module is configured to assist the scheduler in making task allocation decisions according to the task risk assessment information; The step of evaluating the equipment performance degradation degree based on the equipment operation state data and the workpiece quality data comprises: acquiring multi-source machining process signals of equipment operation, processing the multi-source machining process signals of equipment operation to obtain real-time machining response characteristics, acquiring material response characteristic patterns of new alloy materials, the material response characteristic patterns of new alloy materials representing sub-surface micro-defects caused by the interaction between equipment hidden degradation and material characteristics, comparing the real-time machining response characteristics with the material response characteristic patterns of new alloy materials to predict internal defect risks, and comprehensively evaluating the equipment performance degradation degree by combining the internal defect risks with equipment vibration deviation degrees and workpiece surface quality deviation degrees.

[0014] Through the technical scheme, the application provides a system capable of realizing the intelligent scheduling method of the distributed manufacturing resource, the system can effectively integrate various functions through a modular design, provides hardware and software support for actual application, thereby realizing more accurate perception, evaluation and scheduling decision assistance of equipment performance degradation, and significantly improving overall operation efficiency and reliability of the distributed manufacturing system.

[0015] The application realizes more fine and intelligent scheduling management of the distributed manufacturing resource by introducing the equipment performance degradation evaluation and task risk evaluation mechanism. The method overcomes the deficiencies of lagging equipment state information update, lack of effective correlation between quality data and equipment state data, and fragile scheduling decision in the prior art, significantly improves decision accuracy and robustness of the scheduling system, effectively reduces production cost and resource consumption, and improves customer satisfaction and enterprise reputation.

[0016] To enable further understanding of the features and technical contents of the application, please refer to the following detailed description and drawings of the application. However, the drawings provided are only for reference and illustration, and are not used to limit the application. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 A method flow chart of the intelligent scheduling method of the distributed manufacturing resource of the application; Figure 2 A structure schematic diagram of the intelligent scheduling system of the distributed manufacturing resource of the application. DETAILED DESCRIPTION

[0018] The following is to illustrate the embodiments of the application through specific specific embodiments, and the person skilled in the art can understand the advantages and effects of the application from the disclosed contents of the specification. The application can be implemented or applied through other different specific embodiments, and each detail in the specification can be modified and changed based on different viewpoints and applications without departing from the spirit of the application. In addition, the drawings of the application are only simple schematic illustrations, not actual size drawings, and the prior declaration is made. The following embodiments will further illustrate the related technical contents of the application, but the disclosed contents are not used to limit the protection scope of the application.

[0019] The embodiment provides an intelligent scheduling method and system of distributed manufacturing resource, which is combined with Figure 1 and Figure 2 as shown.

[0020] Referring to Figure 1 , an intelligent scheduling method of distributed manufacturing resource, the method comprises the following steps: obtaining equipment running state data and workpiece quality data; Based on the equipment running state data and the workpiece quality data, the degree of equipment performance degradation is evaluated. According to the degree of equipment performance degradation, the equipment capability parameters in the scheduling system are corrected. Before task allocation, based on the degree of equipment performance degradation and task requirements, task risk assessment information is generated. According to the task risk assessment information, the dispatcher is assisted in making task allocation decisions. Among them, the step of evaluating the degree of equipment performance degradation based on the equipment running state data and the workpiece quality data includes: obtaining multi-source machining process signals of equipment running, processing the multi-source machining process signals of equipment running to obtain real-time machining response features, obtaining material response feature patterns of new alloy materials, the material response feature patterns of new alloy materials representing the interaction between equipment implicit degradation and material characteristics leading to subsurface microscopic defects, comparing the real-time machining response features with the material response feature patterns of new alloy materials, predicting internal defect risks, and comprehensively evaluating the degree of equipment performance degradation by combining the internal defect risks with the equipment vibration deviation degree and the workpiece surface quality deviation degree.

[0021] Among them, "equipment running state data" refers to various real-time or quasi-real-time data reflecting the current working condition of the equipment, such as vibration signals, acoustic emission signals, temperature, current, voltage, etc. These data can depict the health status and running stability of the equipment from different dimensions. "Workpiece quality data" refers to the detection data of the finished workpiece in terms of size accuracy, surface roughness, microstructure, etc. These data directly reflect the machining capacity of the equipment and the product quality. "Degree of equipment performance degradation" is a quantitative evaluation of the decline of the current machining capacity of the equipment relative to its design or initial performance, which may manifest as a decline in machining precision, a decrease in efficiency, an increase in failure risk, etc. "Equipment capability parameters in the scheduling system" are the equipment performance indicators on which the scheduling system relies when allocating tasks and scheduling production, such as maximum machining speed, minimum machining time, highest achievable precision level, etc. The accuracy of these parameters directly affects the effectiveness of scheduling decisions. "Task risk assessment information" is a comprehensive quantitative description of various risks (such as delays, unqualified quality, cost increases, etc.) that a specific task may face when executed on a specific equipment.

[0022] In one embodiment, the distributed manufacturing resource intelligent scheduling method first needs to obtain equipment operating state data and workpiece quality data. Equipment operating state data can be obtained by installing various sensors at key parts of the equipment, for example, an acceleration sensor can be deployed to collect vibration signals of the equipment, or an acoustic emission sensor can be deployed to capture acoustic wave signals generated by internal friction, crack propagation, etc. of the equipment. These sensors can continuously and real-time collect data. Workpiece quality data can be obtained by setting online detection equipment at the end of the production line, for example, using an optical detection system to detect the surface quality of the workpiece, or using a three-coordinate measuring machine to measure the dimensional accuracy of the workpiece. After these data are collected, they are usually subjected to preliminary data cleaning and preprocessing to remove noise and outliers, ensuring the accuracy and reliability of the data. For example, a digital filter can be used to filter the collected raw signals to eliminate high-frequency noise interference.

[0023] Subsequently, based on the obtained equipment operating state data and workpiece quality data, the degree of performance degradation of the equipment is evaluated. This can be achieved by establishing an equipment performance model. For example, machine learning algorithms can be used to train a prediction model to output the performance degradation index of the equipment, with equipment operating state data (such as vibration frequency spectrum features, acoustic emission energy) and workpiece quality data (such as dimensional deviation, surface roughness) as inputs. The model can learn the complex nonlinear relationship between equipment operating features and actual processing quality. For example, when there is an increase in energy at a specific frequency in the vibration signal of the equipment, and at the same time accompanied by deterioration of the workpiece surface roughness index, the model can identify that the equipment may have performance degradation such as bearing wear or tool wear.

[0024] Next, according to the evaluated degree of performance degradation of the equipment, the equipment capability parameters in the scheduling system are corrected. For example, if the evaluation result shows that the machining accuracy of a certain equipment has decreased slightly, the "maximum achievable accuracy" parameter of the equipment recorded in the scheduling system should be adjusted accordingly. If the operating efficiency of the equipment decreases, its "unit time capacity" parameter should also be corrected. This correction can be linear or nonlinear, depending on the mapping relationship between the degree of performance degradation and the capability parameters. For example, a correction factor table can be preset, and according to different degradation levels, the equipment capability parameters are adjusted by percentage.

[0025] Before task allocation, task risk assessment information is generated based on the degree of performance degradation of the equipment and the requirements of the task. The requirements of the task can include specific indicators such as machining precision, surface quality, delivery time, etc. For example, if a task requires extremely high machining precision, and the equipment to be allocated is evaluated as having moderate performance degradation, the quality risk of the task performed on the equipment will be evaluated as high. The task risk assessment information can be a comprehensive risk score or a risk report containing multiple dimensions (such as quality risk, delay risk, cost risk). For example, a risk assessment model can be constructed, which inputs the degree of performance degradation of the equipment, the precision requirement of the task, the urgency of the task, etc., and outputs a quantitative task risk index.

[0026] Finally, according to the generated task risk assessment information, the dispatcher is assisted in making task allocation decisions. When making task allocation, the dispatcher no longer relies only on the "normal / fault" state of the equipment, but can obtain deeper insights into the actual processing capacity and potential risks of the equipment. For example, when the scheduling system recommends allocating a high-precision task to a piece of equipment, if the task risk assessment information shows that the equipment has a high quality risk, the dispatcher can choose to allocate the task to another piece of equipment with more stable performance, or take additional quality control measures. This assisted decision-making mechanism enables the dispatcher to make more intelligent and robust decisions, thereby avoiding production delays and quality problems caused by equipment hidden degradation.

[0027] The distributed manufacturing resource intelligent scheduling method proposed in this application significantly improves the depth and breadth of the state perception of the scheduling system for manufacturing resources by introducing real-time and refined evaluation of the degree of performance degradation of the equipment, and combining it with the requirements of the task to generate task risk assessment information.

[0028] Specifically, the traditional scheduling method, when faced with the problem of slight increase in the bearing clearance of the main shaft of the equipment, will misjudge that the equipment is in the "normal" state and allocate tasks based on this misjudgment, since it has not reached the fault alarm threshold. This can lead to high-precision tasks being allocated to equipment whose actual processing capacity has been damaged, ultimately resulting in unqualified product quality and production delays. By obtaining equipment operating state data (such as micro-vibration signals) and workpiece quality data (such as sporadic size deviation), and evaluating the degree of performance degradation of the equipment based on these data, the application can identify hidden problems in the equipment in a timely manner. For example, when the system detects that the energy of a specific frequency band of the equipment vibration signal abnormally increases and is associated with the size deviation of the workpiece, it can be determined that the performance of the equipment is degrading.

[0029] Further, the present application dynamically corrects the equipment capability parameters within the scheduling system according to the evaluated equipment performance degradation. This means that when the processing precision or efficiency of an equipment decreases due to performance degradation, the corresponding parameters within the scheduling system will be adjusted in time, so as to ensure that the subsequent task allocation is based on the true, rather than idealized, capability of the equipment. For example, if the precision evaluation result of a numerical control machine tool shows that it can no longer meet the processing requirements of the highest precision level, its “highest achievable precision” parameter in the scheduling system will be downgraded, so as to avoid allocating tasks with extremely high precision requirements to it.

[0030] In addition, the present application generates task risk assessment information based on the equipment performance degradation and task requirements before task allocation. This enables the scheduler to obtain a more comprehensive risk view when making decisions. For example, for an urgent insert order with strict delivery time requirements and multiple high-precision processes, the present application will not only consider the current load and availability of the equipment, but also combine the equipment performance degradation to predict the delay time or quality unqualified rate that may be generated when the task is executed on the equipment, and quantify it as a production chain transmission effect risk score. This risk assessment information can assist the scheduler to make more intelligent decisions, for example, even if a certain equipment is currently idle, but if its task risk assessment information shows that there is a high quality risk, the scheduler may choose to allocate it to another equipment with more stable performance, or take additional quality control measures, so as to effectively avoid production accidents and customer satisfaction due to equipment hidden problems.

[0031] Specifically, acquiring the multi-source machining process signals of the equipment in operation refers to collecting various physical signals generated by the equipment in the machining process in real time through various sensors deployed on the equipment, such as but not limited to acoustic emission sensors, vibration sensors, force sensors, temperature sensors, etc. These signals can reflect the running state of the equipment and the dynamic changes of the machining process from different dimensions. Among them, processing the multi-source machining process signals of the equipment in operation to obtain real-time machining response features can be understood as pre-processing, feature extraction and feature fusion of the collected original multi-source signals. Pre-processing includes denoising, filtering and other operations; feature extraction can use time domain, frequency domain or time-frequency domain analysis methods to obtain statistical features or spectral features such as energy, amplitude, frequency, kurtosis, margin of the signal; feature fusion aims to integrate the effective information of different types of signals to form real-time machining response features that can fully represent the current machining state. In practical applications, obtaining the material response feature mode of the new alloy material, which represents the subsurface micro-defects caused by the interaction between equipment implicit degradation and material properties, refers to establishing a correlation model between the internal microstructure (such as grain, phase change, dislocation, etc.) and subsurface defects (such as micro-cracks, voids, residual stress, etc.) of the new alloy material and the machining process signal response under different machining conditions and equipment degradation states through previous experiments, simulations or expert knowledge. This mode serves as a benchmark for identifying possible implicit degradation of the equipment when machining new alloy materials. Further, comparing the real-time machining response features with the material response feature mode of the new alloy material to predict internal defect risks refers to matching or difference analysis of the current real-time machining response features and the preset material response feature mode through pattern recognition, machine learning algorithms or statistical analysis methods. When there is a significant deviation between the real-time features and the normal mode, it indicates that the equipment may be causing or exacerbating the subsurface micro-defects of the new alloy material, thereby predicting the potential internal defect risk. Thus, the internal defect risk, the degree of deviation of the equipment vibration and the degree of deviation of the workpiece surface quality are comprehensively evaluated to assess the degree of equipment performance degradation, which refers to weighting fusion or multi-index comprehensive evaluation of the above-mentioned predicted internal defect risk as an important evaluation index, and the traditional equipment vibration deviation (such as abnormal vibration level obtained by analyzing vibration sensor data) and the workpiece surface quality deviation (such as surface roughness, size deviation, etc. data obtained by online optical detection system and three-coordinate measuring machine). Through this comprehensive evaluation, the overall performance degradation of the equipment, including explicit degradation and implicit degradation, can be more comprehensively and accurately reflected.

[0032] The scheme of the present application can capture the micro-dynamic changes of the equipment in the machining process more meticulously by introducing the collection and processing of multi-source machining process signals of the equipment operation. In particular, by obtaining the material response characteristic pattern of the new alloy material and comparing it with the real-time machining response characteristics, the system can identify the subtle degradation of the equipment that is difficult to detect by traditional methods, i.e. the subsurface micro-defects caused by the interaction between the equipment and the new alloy material. The prediction of this hidden defect risk makes up for the shortcomings of relying solely on equipment vibration and workpiece surface quality for evaluation, as many critical defects may first form inside the material or subsurface without obvious external manifestations. Finally, the internal defect risk is integrated with the degree of deviation of the equipment vibration and the degree of deviation of the workpiece surface quality to ensure that the evaluation of the equipment performance degradation is both comprehensive and in-depth, thereby providing a more reliable basis for subsequent scheduling decisions.

[0033] In some preferred embodiments, assuming that a distributed manufacturing system is processing a new titanium alloy material used in the aerospace field. First, through the high-frequency vibration sensors, acoustic emission sensors and cutting force sensors deployed on the main shaft of the machining equipment, the tool and the workpiece clamp, multi-source machining process signals during the machining process are collected in real time. These signals are transmitted to the data processing unit, processed by digital filtering, wavelet transform, etc., to extract real-time machining response characteristics in time domain, frequency domain and time-frequency domain, such as vibration energy spectrum, acoustic emission event rate, cutting force fluctuation amplitude, etc. At the same time, the system has previously established a correlation model between the internal grain structure changes, micro-crack initiation of the new titanium alloy material under different cutting parameters and tool wear states and the above machining response characteristics, i.e. the material response characteristic pattern of the new alloy material, through a large number of experiments and finite element simulations. When the real-time machining response characteristics deviate significantly from this pattern, for example, abnormal peaks of a certain frequency appear in the acoustic emission signal, or the cutting force fluctuation pattern does not match the normal machining pattern, the system predicts the existence of internal defect risks, such as subsurface intergranular cracks or uneven phase transformation. Subsequently, the internal defect risk is quantified as a risk score, and is integrated and weighted with the degree of deviation of the equipment vibration (such as the vibration amplitude of the main shaft bearing exceeding the limit) and the degree of deviation of the workpiece surface quality (such as the surface roughness exceeding the standard detected by the online optical detection system) detected by traditional methods. For example, the weight of the internal defect risk can be set to 0.5, the weight of the vibration deviation degree to 0.3, and the weight of the surface quality deviation degree to 0.2, and finally a comprehensive equipment performance degradation index is calculated. The index will serve as an important basis for the scheduling system to revise the equipment capability parameters and generate task risk assessment information, ensuring the processing quality and reliability of the new titanium alloy workpiece.

[0034] The present application further proposes a step of revising the equipment capability parameters in the scheduling system according to the equipment performance degradation degree, which includes: Obtaining microstructure characteristic data and mechanical property data of the current batch of new materials; Comparing the microstructure characteristic data and mechanical property data of the current batch of new materials with the preset material property benchmark to identify the characteristic deviation degree of the current batch of materials; According to the characteristic deviation degree of the current batch of materials, dynamically adjusting the weight coefficient or adjustment factor for correcting the equipment capability parameters; Using the adjusted weight coefficient or adjustment factor, combined with the equipment performance degradation degree, to correct the equipment capability parameters inside the scheduling system.

[0035] Specifically, obtaining the microstructure characteristic data and mechanical property data of the current batch of new materials means that through professional material detection equipment such as scanning electron microscope, transmission electron microscope, X-ray diffractometer, etc., the microstructure information such as grain size, phase composition, crystal structure, etc. of the material is obtained; at the same time, through tensile testing machine, hardness tester, impact testing machine, etc. The mechanical property indexes such as tensile strength, yield strength, elongation, hardness, toughness, etc. of the material are obtained. These data comprehensively characterize the intrinsic characteristics of the current batch of new materials and are the basis for evaluating their processing performance.

[0036] Among them, the microstructure characteristic data and mechanical property data of the current batch of new materials are compared with the preset material property benchmark to identify the characteristic deviation degree of the current batch of materials. The preset material property benchmark can be understood as the microstructure and mechanical property parameter range of the new material under ideal state or standard batch. Through comparison, the difference between the current batch of materials and the standard benchmark can be quantified, for example, the increase or decrease of grain size, the increase or decrease of hardness, etc. These deviation degrees directly affect the processing behavior of the material and the quality of the final product.

[0037] In practical application, according to the characteristic deviation degree of the current batch of materials, the weight coefficient or adjustment factor for correcting the equipment capability parameters is dynamically adjusted. The weight coefficient or adjustment factor is a multiplier or addition and subtraction amount for quantifying the influence degree of equipment performance degradation degree on equipment capability parameters. For example, if the hardness of the current batch of materials is higher than the benchmark, even if the equipment performance degradation degree is the same, its influence on the processing process may be greater, at this time the weight coefficient can be increased, so that the correction effect of equipment performance degradation degree on equipment capability parameters is more significant; on the contrary, if the material characteristic deviation makes it easier to process, the weight coefficient can be reduced. This dynamic adjustment ensures the flexibility and adaptability of the correction process.

[0038] Thus, the equipment capability parameters in the scheduling system are corrected by the adjusted weight coefficient or adjustment factor in combination with the equipment performance degradation degree. This means that the final equipment capability parameter correction no longer depends only on the equipment degradation state, but also considers the coupling effect of equipment degradation and the deviation of the current batch material characteristics, thereby obtaining a more accurate and practical equipment capability evaluation.

[0039] In some preferred embodiments, assume that a distributed manufacturing system needs to process a batch of new high-strength alloy materials. First, by performing metallographic analysis and tensile test on the batch of alloy materials, the microstructure characteristic data and mechanical property data such as grain size, hardness, tensile strength, etc. are obtained. For example, the test results show that the average grain size of the batch of materials is 10% smaller than the standard benchmark, and the hardness is 5% higher. According to these deviation degrees, the system judges that the processing difficulty of the batch of materials is relatively large. At this time, the scheduling system will dynamically adjust the weight coefficient used to correct the equipment capability parameters. For example, if the equipment performance degradation degree is evaluated as medium, without considering the material characteristics, the equipment capability parameters may be adjusted downward by 5%. But since the hardness of the current batch of materials is higher, the system will adjust the weight coefficient from 1.0 to 1.2, so that the downward adjustment amplitude of the equipment capability parameters becomes 5%*1.2=6%. Finally, the scheduling system will use this equipment capability parameter corrected by the material characteristics to perform task allocation, ensuring that even in the case of equipment degradation to a certain extent, the most suitable equipment can be selected for the high-difficulty material, or the processing parameters can be adjusted to meet the quality requirements.

[0040] Before task allocation, the steps of generating task risk assessment information based on equipment performance degradation degree and task requirements include: Identify the position of the task to be allocated in the product production process and the dependency relationship to the subsequent processes; Predict the delay time or quality unqualified rate that may be generated when the task to be allocated is executed on the equipment; According to the position of the task to be allocated in the product production process and the dependency relationship to the subsequent processes, calculate the influence of the delay time or quality unqualified rate on the start time, completion time of the subsequent processes and product delivery cycle; Quantify the influence as a production chain transmission effect risk score and integrate it into the task risk assessment information.

[0041] Specifically, identifying the position of the to-be-assigned task in the product production process and the dependency relationship with subsequent processes means determining, through analysis of the product process route map or production process map, which stage (e.g., front process, middle process, or rear process) of the production process the current to-be-assigned task is in, and how the completion of the task will affect the start conditions and execution progress of one or more subsequent processes. For example, if a task is a pre-process on the critical path, its delay will directly cause the delay of all subsequent dependent processes.

[0042] Wherein, predicting the delay time or quality unqualified rate that the to-be-assigned task may cause when executed on the equipment can be understood as using a prediction model (e.g., a machine learning model or a statistical model) to estimate the probability and degree of completion time exceeding expectations or product quality not meeting standards that may occur when the task is executed on a specific equipment based on factors such as the degree of performance degradation of the equipment, historical production data, the complexity of the task, and the current load of the equipment. For example, when the degree of performance degradation of the equipment is high, the predicted delay time or quality unqualified rate will increase accordingly.

[0043] In actual application, according to the position of the to-be-assigned task in the product production process and the dependency relationship with subsequent processes, the influence of the delay time or quality unqualified rate on the start time, completion time of subsequent processes and product delivery cycle is calculated, which specifically means taking the predicted delay time or quality unqualified rate as input, combining the production plan and the logical relationship between processes, and evaluating the cumulative impact of these potential problems on the key time nodes (such as the planned start time and completion time of subsequent processes) of the entire production plan and the final product delivery commitment (such as the total delivery cycle) through simulation or chain calculation.

[0044] Further, quantifying the influence as a production chain transmission effect risk score and integrating it into the task risk assessment information means that the above calculated influence on the start time, completion time of subsequent processes and product delivery cycle is converted into a unified, comparable numerical index, i.e., a production chain transmission effect risk score, through a pre-set scoring rule or risk model. This score can reflect the potential destructive power or influence range of the task risk on the entire production chain, and serve as an important part of the task risk assessment information to provide decision basis for the dispatcher.

[0045] In some preferred embodiments, assume that a distributed manufacturing system is producing a new type of aircraft engine blade, and the production process of the blade includes casting, rough machining, finishing, heat treatment, and surface coating, etc. Now there is a to-be-assigned finishing task that needs to be executed on a CNC machine tool with a certain degree of performance degradation.

[0046] First, the system identifies the position of the finishing task in the blade production process, i.e., it is located after rough machining and before heat treatment, and the start of the heat treatment process is strictly dependent on the completion of the finishing task. At the same time, the system identifies that the quality unqualified of the finishing task will directly affect the effectiveness of the subsequent heat treatment and surface coating processes.

[0047] Secondly, based on the degree of performance degradation of the numerical control machine tool (e.g., excessive spindle vibration, high tool wear prediction value) and the strict requirements of the finishing task on precision and surface quality, the system predicts that the finishing task may cause a 2-hour delay when executed on the numerical control machine tool, and there is a 5% probability of causing the blade surface roughness to be unqualified.

[0048] Then, according to the dependence of the finishing task on the subsequent heat treatment process, the system calculates that if the finishing task is delayed for 2 hours, the start time of the heat treatment process will be delayed for 2 hours accordingly; if the quality of the finishing task is unqualified, the entire batch of blades may need to be reworked or scrapped, which will cause the product delivery cycle to be extended by at least one week.

[0049] Finally, the system quantifies the impact of the above 2-hour delay and one-week delivery cycle extension into a production chain transmission effect risk score. For example, a 2-hour delay may be quantified as 50 points, and a one-week delivery cycle extension may be quantified as 200 points. These scores are integrated into the overall risk assessment information of the finishing task. When the scheduler views the task allocation scheme, in addition to seeing the execution risk of the task itself, he or she will also see the potential high risk score of the entire production chain, so as to prioritize the allocation of the task to a more stable and lower-risk device, or take measures in advance (such as preparing a backup device, adjusting the subsequent process plan) to avoid or reduce risks, and ensure the smooth progress of the entire blade production.

[0050] The step of obtaining device operating state data and workpiece quality data includes: Deploying high-sensitivity piezoelectric acceleration sensors and high-precision acoustic emission sensors at key positions of the core machining equipment to collect micro-vibration signals and acoustic emission signals generated during the operation of the equipment; Performing real-time digital filtering processing on the micro-vibration signals and acoustic emission signals; Performing time domain feature extraction and frequency domain feature extraction on the filtered micro-vibration signals; Performing time domain feature extraction and frequency domain feature extraction on the filtered acoustic emission signals; Fusing the extracted vibration features and acoustic emission features to form the device operating state data; An online optical detection system and a three-coordinate measuring machine are deployed at the end of the production line to detect each finished workpiece in real time, and obtain the size deviation and surface roughness data of the workpiece, wherein the size deviation and surface roughness data of the workpiece are workpiece quality data; The workpiece quality data detected in real time is accurately associated with the equipment running state data of the workpiece through the unique identification of the workpiece.

[0051] Specifically, high-sensitivity piezoelectric acceleration sensors and high-precision acoustic emission sensors are deployed at key positions of the core machining equipment, such as the main shaft, the feed shaft, the tool clamping mechanism, etc., to ensure that the micro-vibration signals and acoustic emission signals generated during the operation of the equipment can be captured. These signals are a direct reflection of the dynamic changes of the internal state of the equipment and the machining process. Among them, the micro-vibration signals can represent the state of mechanical wear, looseness or imbalance of the equipment, while the acoustic emission signals can reflect the microscopic events such as material deformation and crack propagation.

[0052] The collected micro-vibration signals and acoustic emission signals are then subjected to real-time digital filtering processing. This processing aims to remove noise interference in the signals, such as environmental noise, electromagnetic interference, etc., to improve the signal-to-noise ratio of the signals and ensure the accuracy of subsequent feature extraction. Digital filtering can use various algorithms, such as Butterworth filter, Chebyshev filter, etc., according to the actual application scenario and signal characteristics.

[0053] The filtered micro-vibration signals and acoustic emission signals are subjected to time domain feature extraction and frequency domain feature extraction respectively. Time domain feature extraction can include root mean square value, peak value, kurtosis, waveform factor, etc., for describing the statistical characteristics of the signals on the time axis. Frequency domain feature extraction obtains the frequency spectrum information of the signals, such as dominant frequency, sideband energy, harmonic component, etc., through Fourier transform, etc., for revealing the energy distribution and periodicity characteristics of the signals on the frequency axis. These features can comprehensively reflect the running state and potential failure modes of the equipment.

[0054] The extracted vibration features and acoustic emission features are then fused to form more comprehensive and robust equipment running state data. Fusion can use various methods, such as feature-level fusion, decision-level fusion, etc. Feature-level fusion means concatenating or weighting combining the feature vectors of different sensors to form a high-dimensional feature vector. Decision-level fusion means classifying or judging the features of different sensors respectively, and then integrating these judgment results. Through fusion, the complementarity of information from different sensors can be fully utilized to improve the accuracy and reliability of the perception of the equipment running state.

[0055] At the end of the production line, an online optical detection system and a three-coordinate measuring machine are deployed to detect each finished workpiece in real time. The online optical detection system can quickly obtain the surface topography, size profile, and other information of the workpiece, such as identifying surface defects and measuring key dimensions through machine vision technology. The three-coordinate measuring machine provides high-precision size measurement to obtain size deviation and surface roughness data of the workpiece. These data are defined as workpiece quality data, which are important basis for evaluating the quality of the machining process and the processing capacity of the equipment.

[0056] To realize the correlation analysis between the equipment running state and the workpiece quality, the real-time detected workpiece quality data is accurately correlated with the equipment running state data for machining the workpiece through the unique identification of the workpiece. The unique identification of the workpiece can be a two-dimensional code, an RFID tag, or a production batch number, etc., ensuring that the quality data of each workpiece can be accurately traced back to its corresponding machining equipment and machining process. This accurate correlation provides a data basis for subsequent equipment performance degradation evaluation.

[0057] The scheme of the present application can comprehensively and real-timely collect the micro-vibration signals and acoustic emission signals in the equipment running process by deploying high-sensitivity piezoelectric acceleration sensors and high-precision acoustic emission sensors at key positions of the core machining equipment. These signals are effectively converted into vibration features and acoustic emission features that can represent the equipment health status and the dynamic of the machining process after real-time digital filtering processing, time domain feature extraction, and frequency domain feature extraction. By fusing these multi-source features, more comprehensive and accurate equipment running state data can be formed, overcoming the limitations of insufficient information of a single sensor. At the same time, the online optical detection system and the three-coordinate measuring machine are deployed at the end of the production line to realize real-time and high-precision quality detection of each finished workpiece, and key size deviation and surface roughness data are obtained. Finally, the equipment running state data is accurately correlated with the workpiece quality data through the unique identification of the workpiece, and a direct mapping relationship between the equipment state and the product quality is established, providing a solid data basis for subsequent equipment performance degradation evaluation and intelligent scheduling.

[0058] The present application further proposes a step of fusing the extracted vibration features and acoustic emission features to form the equipment running state data, comprising: obtaining current working mode information of the equipment; according to the current working mode information of the equipment, obtaining vibration feature fusion weights and acoustic emission feature fusion weights corresponding to the current working mode from a preset working mode and fusion weight mapping rule; using the vibration feature fusion weights and the acoustic emission feature fusion weights to weight and fuse the extracted vibration features and acoustic emission features to form the equipment running state data.

[0059] Specifically, obtaining the current working mode information of the equipment refers to obtaining the current running state or processing stage of the equipment in real time through the controller of the equipment itself, sensor data analysis or operator input, etc. For example, the working mode information can include rough machining, finishing, no load, load, standby, etc. Among them, the preset working mode and the fusion weight mapping rule can be understood as a database or lookup table stored in the scheduling system or the equipment control system. This rule predefines the relative importance or contribution of vibration features and acoustic emission features in forming equipment running state data under different working modes. For example, in some working modes, vibration features may be more sensitive to equipment performance degradation, so their fusion weight will be higher; while in other working modes, acoustic emission features may provide more critical information, and their fusion weight will be correspondingly increased. These weight coefficients or adjustment factors can be determined by historical data analysis, expert experience or machine learning model training, etc. In practical application, the vibration feature fusion weight and the acoustic emission feature fusion weight are numerical parameters used for weighted processing of the extracted vibration features and acoustic emission features. The setting of these weight coefficients aims to ensure that the fused equipment running state data can more accurately and comprehensively reflect the real health status of the equipment under different working modes. Further, the weighted fusion of the extracted vibration features and acoustic emission features to form the equipment running state data refers to multiplying the vibration features and acoustic emission features after time domain feature extraction and frequency domain feature extraction by their corresponding vibration feature fusion weight and acoustic emission feature fusion weight, respectively, and then combining or superimposing the weighted features to generate a comprehensive equipment running state data. This weighted fusion method can dynamically adjust the contribution of different types of features according to the current working mode of the equipment, making the final equipment running state data more representative and accurate.

[0060] In some preferred embodiments, assuming that a numerical control machine tool is executing a machining task, there are three typical working modes: "rough machining", "fine machining", and "no-load". When the device is in the "rough machining" mode, it is more prone to generate larger mechanical vibrations due to large cutting amount and strong cutting force, and at this time the vibration characteristics are more sensitive to performance degradation such as wear and looseness. Therefore, in the preset working mode and fusion weight mapping rule, for the "rough machining" mode, the vibration feature fusion weight is set to 0.7, and the acoustic emission feature fusion weight is set to 0.3. When the device switches to the "fine machining" mode, the cutting amount is reduced, and the machining precision and surface quality are required to be higher, at this time the acoustic emission signal may be more able to reflect the micro wear of the tool, the generation of internal defects of the material, etc. Therefore, for the "fine machining" mode, the vibration feature fusion weight is set to 0.4, and the acoustic emission feature fusion weight is set to 0.6. When the device is in the "no-load" mode, the main concern is the stability of the idle running of the device, the wear of the bearing, etc., at this time the characteristics of the vibration and acoustic emission signals may be significantly different from those under load. For the "no-load" mode, the vibration feature fusion weight and the acoustic emission feature fusion weight may be set to 0.5 and 0.5, or more fine adjustment according to the actual situation. In this way, the system can dynamically obtain the corresponding fusion weights from the preset rules according to the current working mode information of the device, and use these weights to weight and fuse the real-time extracted vibration features and acoustic emission features. Thus, under different working modes, the device running state data formed can more accurately reflect the key health indicators of the device under that mode, avoiding misjudgment or information omission that may be caused by a single weight strategy, thereby providing a more reliable basis for subsequent intelligent scheduling decisions.

[0061] The step of fusing the extracted vibration features and acoustic emission features to form the device running state data includes: When the device working mode is switched, the working mode information before switching and the working mode information after switching are obtained; According to the working mode information before switching and the working mode information after switching, a transition phase of the working mode switching is determined; In the transition phase of the working mode switching, a time decay based weighting fusion strategy is used to interpolate the fusion weight corresponding to the working mode before switching and the fusion weight corresponding to the working mode after switching to obtain a transition fusion weight; The transition fusion weight is used to weight and fuse the extracted vibration features and acoustic emission features to form the device running state data.

[0062] Specifically, at the time of switching the working mode of the machine tool, the system is configured to monitor the working state of the machine tool in real time, and once the change of the working mode is detected, the working mode information before the switching and the working mode information after the switching are obtained. For example, when the machine tool is switched from the "rough machining mode" to the "fine machining mode", the system records the "rough machining mode" as the working mode information before the switching, and the "fine machining mode" as the working mode information after the switching. These working mode information can include mode identifiers, preset machining parameter ranges, expected load characteristics, etc.

[0063] Further, according to the obtained working mode information before the switching and the working mode information after the switching, the system is used to determine the transition phase of the working mode switching. The transition phase refers to the time interval during which the machine tool gradually evolves from one stable working state to another stable working state. For example, the start point and the end point of the transition phase can be defined by analyzing the instruction sequence of the machine tool control system, the trend of the internal sensor data of the machine tool (such as spindle speed, feed speed, tool load, etc.). The length of the transition phase can be preset or dynamically adjusted according to the specific type of the machine tool, the difference between the modes of the switching, and the empirical data.

[0064] During the transition phase of the working mode switching, in order to more accurately reflect the running state of the machine tool, the application adopts a time-decay-based weighted fusion strategy. This strategy aims to interpolate the fusion weight corresponding to the working mode before the switching and the fusion weight corresponding to the working mode after the switching, so as to obtain the transition fusion weight. Specifically, as time moves from the start point to the end point of the transition phase, the influence of the fusion weight of the working mode before the switching gradually weakens, while the influence of the fusion weight of the working mode after the switching gradually strengthens. For example, linear interpolation, exponential decay or S-shaped curve functions can be used to calculate the transition fusion weight at different time points. These fusion weights include vibration feature fusion weights and acoustic emission feature fusion weights, which are respectively used to weight the extracted vibration features and acoustic emission features.

[0065] Finally, using the obtained transition fusion weight, the extracted vibration features and acoustic emission features are weighted and fused, so as to form more accurate machine tool running state data during the transition phase of the working mode switching. This weighted fusion process is similar to the fusion method in the above embodiments, but the key lies in the dynamic adjustment of the fusion weight to adapt to the dynamic characteristics of the machine tool during the mode switching process.

[0066] In some preferred embodiments, it is assumed that a numerical control milling machine needs to be switched from the "rough milling mode" to the "fine milling mode" when performing a machining task of a complex part.

[0067] Firstly, when the system detects the switching instruction from the "rough milling mode" to the "fine milling mode", the working mode information before switching (rough milling mode) and the working mode information after switching (fine milling mode) are immediately obtained.

[0068] Then, the system determines the transition stage of this mode switching to be 10 seconds according to the preset rules or by analyzing the device parameters (such as the change curves of spindle speed, feed speed, tool load, etc.) in real time.

[0069] In the next 10 seconds, the system will continuously monitor the micro-vibration signals and acoustic emission signals generated during the operation of the equipment, and extract the corresponding vibration features and acoustic emission features. At the same time, the system will dynamically calculate the transition fusion weight according to the time decay-based weighted fusion strategy. For example, at the beginning of the transition stage (t=0s), the fusion weight may be more biased towards the weight of the rough milling mode; at the end of the transition stage (t=10s), the fusion weight is completely transitioned to the weight of the fine milling mode. At the intermediate time (such as t=5s), the fusion weight will be an interpolation result of the rough milling mode weight and the fine milling mode weight, for example, a linear interpolation can be used, that is, the transition fusion weight=(1-t / 10)*rough milling mode weight+(t / 10)*fine milling mode weight.

[0070] Finally, the system uses these dynamically calculated transition fusion weights to perform weighted fusion on the real-time extracted vibration features and acoustic emission features, thereby generating more accurate equipment operation state data during the mode switching transition stage. These data are then used to evaluate the degree of equipment performance degradation, ensuring that even in dynamic switching scenarios, the equipment health status can be accurately judged, and then assisting the dispatcher to make more optimal task allocation decisions.

[0071] The present application further proposes that before task allocation, the step of generating task risk assessment information based on the degree of equipment performance degradation and task requirements includes: Obtaining specific index threshold values of the product microstructure or surface topography in the task requirements; According to the degree of equipment performance degradation and the specific index threshold values, predicting the risk level of subsurface defect types and internal stress concentration generated during the machining process; Obtaining the correlation rules between the long-term reliability or specific functions of the product and the risk level of subsurface defect types and internal stress concentration; According to the correlation rules, quantifying the influence of the risk level of subsurface defect types and internal stress concentration on the long-term reliability or specific functions of the product; The quantified influence is taken as the hidden defect risk score and integrated into the task risk assessment information.

[0072] Specifically, obtaining specific indicator thresholds of product microstructure or surface topography in task requirements means that when receiving a task to be allocated, the system will analyze the specific quality requirements of the final product at the micro level (such as grain size, phase composition, inclusion distribution) or the macro surface level (such as surface roughness, waviness, residual stress). These requirements are usually given in the form of quantitative indicator thresholds, for example, the grain size is required to be less than 10 microns, or the surface roughness Ra value is required to be less than 0.2 microns. The purpose is to provide a clear quality benchmark for subsequent implicit defect prediction and risk assessment.

[0073] Among them, according to the degree of equipment performance degradation and the specific indicator threshold, the risk level of subsurface defect type and internal stress concentration generated in the machining process can be understood as the system using the evaluated degree of equipment performance degradation (such as spindle vibration anomaly, tool wear aggravation, etc.) and the specific indicator threshold of task requirements for comprehensive analysis. For example, when the equipment degrades to a certain extent, combined with the strict requirements of the task on the surface topography, the system can predict that microcracks, pores, non-metallic inclusions and other subsurface defects may be generated in the machining process, and evaluate the risk level of internal stress concentration caused by these defects. The purpose is to identify potential, naked eye or conventional detection difficult to find quality problems in advance.

[0074] In practical application, obtaining the correlation rules between product long-term reliability or specific function and subsurface defect type and internal stress concentration risk level specifically refers to establishing a knowledge base or model that contains historical data, expert experience, simulation results, etc., which is used to describe how different types of subsurface defects and internal stress concentration risk levels affect the long-term service performance (such as fatigue life, creep performance, corrosion resistance) or specific function (such as electrical conductivity, optical performance) of the product. For example, a certain type of microcrack may cause the product fatigue life to decrease by 20%, and a certain degree of internal stress concentration may increase the failure probability of the product under a certain load. The purpose is to provide a scientific basis for quantifying the impact of implicit defects on the product.

[0075] Further, according to the correlation rules, quantifying the impact of subsurface defect type and internal stress concentration risk level on product long-term reliability or specific function means that the system converts the predicted subsurface defect type and internal stress concentration risk level into a specific impact value on product long-term reliability or specific function according to the above correlation rules. For example, if a certain subsurface defect is predicted, and the defect will cause the product fatigue life to decrease according to the correlation rules, then the decrease is quantified as a specific impact value. The purpose is to convert the abstract defect risk into a measurable specific impact on product performance.

[0076] Thus, the impact of quantification is incorporated into the task risk assessment information as an implicit defect risk score, which means that the quantified impact value is integrated into the overall task risk assessment information as an independent or weighted score. This implicit defect risk score can reflect the risk that the task may cause the product to produce an implicit defect and affect its long-term performance due to the interaction of equipment degradation and material characteristics when the task is executed on a specific device. The purpose is to make the task risk assessment more comprehensive and in-depth, not only considering explicit risks, but also considering implicit risks.

[0077] In some preferred embodiments, assume that a distributed manufacturing system needs to process a batch of new high-temperature alloy materials for aircraft engine blades. The task requires the blades to have extremely high fatigue life and creep resistance, and has strict index thresholds for the microstructure of the material and the surface residual stress.

[0078] First, the system will obtain the specific index thresholds for the microstructure of the high-temperature alloy blade in the task requirements, such as the average grain size being less than 50 microns and the surface residual compressive stress being greater than -300 MPa.

[0079] Second, combined with the performance degradation of a certain processing equipment (for example, a five-axis milling machine), such as abnormal micro-vibration caused by spindle bearing wear and high tool wear, the system will predict the possible sub-surface micro-crack types (for example, micro-cracks distributed along the grain boundary) and internal tensile stress concentration risk levels when processing on this equipment. For example, it is predicted that moderate sub-surface micro-cracks and high-risk internal tensile stress concentration may occur in the blade root area.

[0080] Next, the system will obtain the association rules between product long-term reliability (for example, fatigue life) and sub-surface micro-crack types and internal stress concentration risk levels from a pre-set knowledge base. For example, the rules may indicate that a certain type of sub-surface micro-crack and high-risk internal tensile stress concentration may reduce the fatigue life of the high-temperature alloy blade by 15% to 20%.

[0081] Then, according to these association rules, the system will quantify the predicted sub-surface micro-cracks and internal tensile stress concentration risks into specific values of impact on blade fatigue life, for example, quantified as a 18% reduction in fatigue life.

[0082] Finally, the quantified impact (i.e., 18% reduction in fatigue life) is incorporated into the overall risk assessment information for the task as a hidden defect risk score. When viewing the risk assessment information for the task, the scheduler will see not only the regular production cycle delay risk or surface quality nonconformance rate risk, but also a clear hidden defect risk score indicating that processing the task on this equipment can lead to a significant reduction in the fatigue life of the product. Based on this, the scheduler can make more informed decisions, such as assigning the task to a more capable equipment, or adjusting the processing parameters to reduce the risk, or even considering adding an additional non-destructive testing step, thereby ensuring the long-term reliability and safety of the aircraft engine blades.

[0083] The application further proposes that the step of incorporating the quantified impact as a hidden defect risk score into the task risk assessment information includes: obtaining hidden defect type information contained in the quantified impact; According to the hidden defect type information, obtaining the risk presentation granularity or dimension corresponding to the hidden defect type from the preset defect type and risk presentation rule; According to the risk presentation granularity or dimension, the quantified impact is converted to generate a hidden defect risk score with a specific granularity or dimension; Incorporate the hidden defect risk score with a specific granularity or dimension into the task risk assessment information.

[0084] Specifically, obtaining hidden defect type information contained in the quantified impact means identifying and extracting the specific hidden defect type that causes the impact from the calculation result of the quantified impact. These hidden defect types can include but are not limited to microcracks, pores, inclusions, grain boundary segregation, residual stress, etc., and the purpose is to clarify the source and nature of the risk.

[0085] Among them, according to the hidden defect type information, the risk presentation granularity or dimension corresponding to the hidden defect type is obtained from the preset defect type and risk presentation rule, which means that the system maintains a knowledge base or database containing multiple defect types and their corresponding risk presentation methods. For example, for microcracks, it may be necessary to present its risk in the form of a spatial distribution map to indicate the specific location where the crack may occur; for internal stress concentration, it may be necessary to present its risk evolution trend in the form of a time series chart to predict the critical point of stress accumulation. Risk presentation granularity or dimension can include but is not limited to spatial dimension (such as defect location, distribution density), time dimension (such as defect evolution rate, duration), severity dimension (such as defect size, number), impact range dimension (such as impact on subsequent processes) and the like. The purpose is to provide guidance for subsequent risk score conversion and ensure that risk information can be understood and utilized in the most effective way.

[0086] In practical applications, converting the quantified impact according to the risk presentation granularity or dimension to generate a hidden defect risk score with a specific granularity or dimension refers to reconstructing, refining, or aggregating the original quantified impact data according to the obtained risk presentation granularity or dimension. For example, if the risk presentation dimension is spatial distribution, the quantified impact can be decomposed into risk values for different regions; if the risk presentation granularity is severity level, the quantified impact can be mapped to predefined risk levels (such as low, medium, and high). This conversion can be achieved using various data processing and visualization techniques, and the purpose is to convert the abstract quantified impact into a specific, intuitive, and easy-to-understand risk score form for decision makers.

[0087] Further, integrating the hidden defect risk score with a specific granularity or dimension into task risk assessment information refers to presenting the converted and structured hidden defect risk score as a component of task risk assessment information to the dispatcher together with other risk assessment indicators (such as delay time, quality failure rate, etc.). This can be displayed in various forms such as charts, heat maps, risk level indicators, etc. on the dispatching interface, and the purpose is to provide a comprehensive, multi-dimensional risk view for the dispatcher to assist in making more intelligent task allocation decisions.

[0088] In some preferred embodiments, assuming that a batch of new alloy materials are being processed, two main hidden defect risks are predicted by the above method: one is the subsurface micro-crack risk, and the other is the internal residual stress concentration risk.

[0089] First, the system obtains the hidden defect type information contained in the quantified impact, i.e., "subsurface micro-crack" and "internal residual stress concentration".

[0090] Next, according to the preset defect type and risk presentation rule, the system identifies that for "subsurface micro-crack", the corresponding risk presentation granularity or dimension is "spatial distribution and size", because the location and size of micro-cracks have a significant impact on subsequent fatigue life; and for "internal residual stress concentration", the corresponding risk presentation granularity or dimension is "stress gradient and regional range", because the gradient and influence range of stress concentration determine the threat to structural integrity.

[0091] According to these risk presentation granularities or dimensions, the system converts the original quantified impact. For example, for the micro-crack risk, a three-dimensional risk map containing potential crack location coordinates, predicted crack length and depth is generated as a hidden defect risk score; for the residual stress concentration risk, a two-dimensional heat map showing the stress concentration area, maximum stress value and stress gradient change is generated as a hidden defect risk score.

[0092] Finally, these hidden defect risk scores with specific granularity or dimension are integrated into the task risk assessment information and presented to the scheduler in a visualized manner. The scheduler can clearly see which areas may have micro-cracks and which parts have a higher risk of residual stress concentration. Based on these refined information, the scheduler can decide to assign the task to a device with more accurate processing capabilities or equipped with online stress monitoring functions, or adjust the processing path to avoid high-risk areas, or even arrange additional non-destructive testing procedures, thereby effectively avoiding product failure caused by hidden defects and ensuring the processing quality and product reliability of new alloy materials.

[0093] Reference Figure 2 The present application also proposes a distributed manufacturing resource intelligent scheduling system for implementing the distributed manufacturing resource intelligent scheduling method. The system comprises: an acquisition module configured to acquire device operating state data and workpiece quality data; a performance evaluation module configured to evaluate the degree of performance degradation of the device based on the device operating state data and the workpiece quality data; a correction module configured to correct the device capability parameters within the scheduling system according to the degree of performance degradation of the device; a risk assessment module configured to generate task risk assessment information based on the degree of performance degradation of the device and the task requirements before task allocation; an assistance module configured to assist the scheduler in making task allocation decisions based on the task risk assessment information; The step of evaluating the degree of performance degradation of the device based on the device operating state data and the workpiece quality data comprises: acquiring multi-source machining process signals of the device operation, processing the multi-source machining process signals of the device operation to obtain real-time machining response characteristics, acquiring material response characteristic patterns of the new alloy material, the material response characteristic patterns of the new alloy material representing the sub-surface microscopic defects caused by the interaction between device hidden degradation and material characteristics, comparing the real-time machining response characteristics with the material response characteristic patterns of the new alloy material to predict internal defect risks, and comprehensively evaluating the degree of performance degradation of the device by combining the internal defect risks with the device vibration deviation degree and the workpiece surface quality deviation degree.

[0094] The acquisition module is configured to collect necessary raw data from the distributed manufacturing environment. Specifically, this module can integrate various sensor interfaces, data acquisition units, and data exchange interfaces with existing manufacturing execution systems or enterprise resource planning systems to acquire real-time device operating state data and workpiece quality data after machining. The device operating state data can include but is not limited to device vibration, temperature, current, power, etc. While the workpiece quality data can include size deviation, surface roughness, internal defects, etc.

[0095] The performance evaluation module is configured to receive the equipment operation state data and workpiece quality data provided by the acquisition module, and quantitatively evaluate the degree of performance degradation of the equipment based on these data. This module can use advanced data analysis algorithms, such as machine learning models, statistical process control methods, or degradation prediction algorithms based on physical models, to identify the performance degradation trend of the equipment in long-term operation or specific working conditions. The evaluation results are usually expressed in numerical or rating form to reflect the health status and potential failure risk of the equipment.

[0096] The correction module is configured to dynamically correct the equipment capability parameters maintained within the scheduling system according to the degree of equipment performance degradation output by the performance evaluation module. In actual manufacturing scheduling, the theoretical capability parameters of the equipment may not accurately reflect its real performance in actual operation, especially when the equipment has performance degradation. By introducing the degradation degree as a correction factor, this correction module can more accurately adjust parameters such as available capacity, processing precision, and reliability of the equipment, ensuring the accuracy and effectiveness of scheduling decisions.

[0097] The risk assessment module is configured to consider the degree of performance degradation of the equipment and the specific requirements of the tasks to be assigned before task allocation, and generate task risk assessment information. This module can analyze the requirements of the task on equipment precision, stability, reliability, etc., and combine the current degradation state of the equipment to predict the risks that the task may face when executed on a specific equipment, such as unqualified processing quality, production delay, or equipment failure, etc. Risk assessment information can include risk level, potential loss estimation, and probability of risk occurrence, etc.

[0098] The auxiliary module is configured to receive the task risk assessment information generated by the risk assessment module, and based on this, assist the scheduler in making task allocation decisions. This module can provide an intuitive risk visualization interface to show the scheduler the risk level under different task allocation schemes, and can recommend the optimal task allocation scheme according to the preset scheduling strategy or optimization goal. Through this module, the scheduler can more comprehensively understand the potential risks of task allocation, making more intelligent and robust decisions.

[0099] The scheme of the present application realizes real-time perception of manufacturing resource state, accurate evaluation of performance, dynamic correction of scheduling parameters, and intelligent early warning of task risk by embodying each logical step in the distributed manufacturing resource intelligent scheduling method into mutually cooperative system modules.

[0100] The above disclosed content is only the preferred feasible embodiment of the present application, and does not limit the protection scope of the present application, so any equivalent technical changes made according to the content of the present application specification and drawings are included in the protection scope of the present application, and in addition, the elements can be updated as technology develops.

Claims

1. A distributed intelligent scheduling method for manufacturing resources, characterized in that, The method includes the following steps: Acquire equipment operating status data and workpiece quality data; Based on equipment operating status data and workpiece quality data, assess the degree of equipment performance degradation; Adjust the equipment capability parameters within the scheduling system according to the degree of equipment performance degradation; Before task assignment, task risk assessment information is generated based on the degree of equipment performance degradation and task requirements; Based on task risk assessment information, assist the scheduler in making task allocation decisions; The steps for assessing the degree of equipment performance degradation based on equipment operating status data and workpiece quality data include: acquiring multi-source processing signals of equipment operation; processing the multi-source processing signals of equipment operation to obtain real-time processing response characteristics; acquiring the material response characteristic pattern of the new alloy material; the material response characteristic pattern of the new alloy material characterizes the subsurface micro-defects caused by the interaction between equipment latent degradation and material properties; comparing the real-time processing response characteristics with the material response characteristic pattern of the new alloy material to predict the risk of internal defects; and comprehensively assessing the degree of equipment performance degradation by combining the risk of internal defects with the degree of equipment vibration deviation and the degree of workpiece surface quality deviation.

2. The distributed manufacturing resource intelligent scheduling method as described in claim 1, characterized in that, The steps for adjusting the equipment capability parameters within the scheduling system based on the degree of equipment performance degradation include: To obtain microstructural characteristics and mechanical properties data of the current batch of novel materials; The microstructure characteristics and mechanical properties of the current batch of new materials are compared with the preset material property benchmarks to identify the degree of deviation of the properties of the current batch of materials. Based on the degree of deviation of the characteristics of the current batch of materials, dynamically adjust the weighting coefficients or adjustment factors used to correct the equipment capacity parameters; By using the adjusted weighting coefficients or adjustment factors, and in conjunction with the degree of equipment performance degradation, the equipment capability parameters within the scheduling system are corrected.

3. The distributed manufacturing resource intelligent scheduling method as described in claim 1, characterized in that, Before task assignment, the steps for generating task risk assessment information based on the degree of equipment performance degradation and task requirements include: Identify the location of the tasks to be assigned in the product manufacturing process and their dependencies on subsequent processes; Predict the potential delays or quality non-conformities that may occur when assigned tasks are executed on the equipment; Based on the position of the task to be assigned in the product production process and its dependence on subsequent processes, calculate the impact of delay time or quality non-conformity rate on the start time and completion time of subsequent processes and product delivery cycle. The impact will be quantified as a risk score for the transmission effect along the production chain and incorporated into the task risk assessment information.

4. The distributed manufacturing resource intelligent scheduling method as described in claim 1, characterized in that, The steps for obtaining equipment operating status data and workpiece quality data include: High-sensitivity piezoelectric accelerometers and high-precision acoustic emission sensors are deployed in key parts of the core processing equipment to collect minute vibration signals and acoustic emission signals generated during equipment operation; Real-time digital filtering processing of minute vibration signals and acoustic emission signals; Time-domain and frequency-domain features are extracted from filtered minute vibration signals. Time-domain and frequency-domain features are extracted from the filtered acoustic emission signal. The extracted vibration features and acoustic emission features are fused to form equipment operating status data; An online optical inspection system and a coordinate measuring machine are deployed at the end of the production line to perform real-time inspection on each finished workpiece and obtain the workpiece's dimensional deviation and surface roughness data. The obtained dimensional deviation and surface roughness data are the workpiece quality data. By using a unique workpiece identifier, the real-time detected workpiece quality data can be accurately correlated with the operating status data of the equipment that processes the workpiece.

5. The distributed manufacturing resource intelligent scheduling method as described in claim 4, characterized in that, The steps for fusing the extracted vibration features and acoustic emission features to form equipment operating status data include: Obtain the current operating mode information of the device; Based on the current working mode information of the equipment, obtain the vibration feature fusion weight and acoustic emission feature fusion weight corresponding to the current working mode from the preset working mode and fusion weight mapping rules; By using vibration feature fusion weights and acoustic emission feature fusion weights, the extracted vibration features and acoustic emission features are weighted and fused to form equipment operating status data.

6. The distributed manufacturing resource intelligent scheduling method as described in claim 5, characterized in that, The steps for fusing the extracted vibration features and acoustic emission features to form equipment operating status data include: When switching the device's operating mode, obtain the operating mode information before the switch and the operating mode information after the switch; Based on the working mode information before and after the switch, determine the transition phase of the working mode switch. During the transition phase of switching working modes, a time decay-based weighted fusion strategy is adopted to interpolate the fusion weights corresponding to the working modes before and after the switch to obtain the transition fusion weights. By using transition fusion weights, the extracted vibration features and acoustic emission features are weighted and fused to form equipment operating status data.

7. The distributed manufacturing resource intelligent scheduling method as described in claim 1, characterized in that, Before task assignment, the steps for generating task risk assessment information based on the degree of equipment performance degradation and task requirements include: Obtain the specific threshold values ​​for the product's microstructure or surface morphology as required by the task; Based on the degree of equipment performance degradation and specific index thresholds, predict the types of subsurface defects and the risk level of internal stress concentration generated during processing; Obtain the correlation rules between long-term product reliability or specific functions and subsurface defect types and internal stress concentration risk levels; Based on association rules, the impact of subsurface defect types and internal stress concentration risk levels on the long-term reliability or specific functions of the product is quantified. The quantified impact is used as a latent defect risk score and integrated into the task risk assessment information.

8. The distributed manufacturing resource intelligent scheduling method as described in claim 7, characterized in that, The steps to incorporate quantified impact as a latent defect risk score into task risk assessment information include: Obtain information on the types of latent defects contained in the quantified impact; Based on the information on the type of latent defect, obtain the risk presentation granularity or dimension corresponding to the latent defect type from the preset defect type and risk presentation rules; Based on the granularity or dimension of risk presentation, the quantitative impact is transformed to generate a latent defect risk score with a specific granularity or dimension; Incorporate latent defect risk scores with specific granularity or dimensions into the task risk assessment information.

9. A distributed manufacturing resource intelligent scheduling system, applied to the distributed manufacturing resource intelligent scheduling method as described in claim 1, characterized in that, The system includes: The acquisition module is used to acquire equipment operating status data and workpiece quality data; The performance evaluation module assesses the degree of equipment performance degradation based on equipment operating status data and workpiece quality data. The correction module adjusts the equipment capability parameters within the scheduling system based on the degree of equipment performance degradation. The risk assessment module generates task risk assessment information based on the degree of equipment performance degradation and task requirements before task allocation; The auxiliary module assists the scheduler in making task allocation decisions based on task risk assessment information. The steps for assessing the degree of equipment performance degradation based on equipment operating status data and workpiece quality data include: acquiring multi-source processing signals of equipment operation; processing the multi-source processing signals of equipment operation to obtain real-time processing response characteristics; acquiring the material response characteristic pattern of the new alloy material; the material response characteristic pattern of the new alloy material characterizes the subsurface micro-defects caused by the interaction between equipment latent degradation and material properties; comparing the real-time processing response characteristics with the material response characteristic pattern of the new alloy material to predict the risk of internal defects; and comprehensively assessing the degree of equipment performance degradation by combining the risk of internal defects with the degree of equipment vibration deviation and the degree of workpiece surface quality deviation.

Images