A digital intelligent factory dynamic planning and scheduling system

Abstract

The application relates to the technical field of data management, and particularly discloses a dynamic planning and scheduling system for a digital intelligent factory, which comprises a real-time state identification module, a degradation generation module, a health degradation module, a plan preset module, a reallocation module and an instruction execution module. The real-time state tensor of the factory is generated by combining the real-time working condition parameters of the equipment, the real-time inventory level of the materials and the order delivery priority. The characteristic degradation curve of the key moving parts of the equipment is obtained by continuously analyzing the equipment working condition parameters in the tensor in the time domain. The equipment sub-health warning time stamp is predicted according to the inflection point of the curve, and is collected into a dynamic constraint matrix. The time and space resource conflict node of the overlapping period of the process and the equipment is identified by combining the matrix and the order priority. The adaptive replacement equipment for the conflict process is screened, and a reallocation scheme is formed. Finally, the scheme is disassembled into three types of instructions, and is synchronously issued to the corresponding execution end. The application can improve the dynamic planning and scheduling efficiency of the digital intelligent factory.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a dynamic planning and scheduling system for intelligent factories. Background Technology

[0002] In the production and operation of smart factories, planning and scheduling, as a core link in production control, plays a crucial supporting role in production efficiency and order delivery efficiency. However, existing factory planning and scheduling systems mostly rely on static equipment parameters, predetermined process routes, and fixed order information for production planning. They cannot collect and integrate dynamic production data such as real-time equipment operating parameters, real-time material inventory levels, and order delivery priorities. They also lack a monitoring and prediction mechanism for equipment health status, resulting in a significant disconnect between the scheduling plan and the actual production status of the factory. It is difficult to predict production resource problems caused by changes in equipment status in advance, which can easily lead to a mismatch between process processing and equipment operating status.

[0003] Existing scheduling systems lack a systematic approach to identifying and handling conflict nodes when dealing with production resource conflicts. They cannot accurately pinpoint the spatiotemporal resource conflicts arising from the overlap between equipment sub-health warning times and scheduled process times. Furthermore, they fail to establish scientific priority ranking rules for conflict nodes in conjunction with order delivery requirements. In the resource reallocation phase of conflicting processes, the selection criteria for alternative equipment are too simplistic, failing to comprehensively consider factors such as equipment load rate and dynamic operating constraints. Moreover, the issuance of instructions after scheduling adjustments lacks coordination, making it impossible to achieve synchronous updates of materials, processes, and work instructions. This results in low efficiency in the dynamic adjustment of scheduling schemes, making it difficult to adapt to the flexible and efficient production needs of smart factories. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides a dynamic planning and scheduling system for intelligent factories, characterized in that the system includes a real-time status identification module, a degradation generation module, a health degradation module, a plan preset module, a reallocation module, and an instruction execution module, wherein:

[0005] The real-time status recognition module is used to generate a real-time status tensor of the factory based on the real-time operating parameters of the equipment, the real-time inventory level of materials, and the order delivery priority in the factory.

[0006] The degradation generation module is used to perform continuous time-domain analysis on the real-time operating parameters of the equipment in the real-time state tensor to obtain the characteristic degradation curves of the key moving parts of the equipment in the factory.

[0007] The health degradation module is used to predict the warning timestamps of equipment in the factory entering a sub-healthy state based on the inflection points of the characteristic degradation curves, and to aggregate the warning timestamps into a dynamic constraint matrix.

[0008] The planning preset module is used to identify spatiotemporal resource conflict nodes in the factory during the planning execution cycle, based on the order delivery priority in the dynamic constraint matrix and the real-time state tensor, caused by the overlap between the precursor timestamp and the predetermined process processing time.

[0009] The reallocation module is used to select alternative equipment with the same processing function and whose current load rate is lower than a preset elastic threshold for the conflicting processes of the factory based on the spatiotemporal resource conflict nodes, so as to obtain the reallocation scheme of the factory.

[0010] The instruction execution module is used to decompose the redistribution scheme into material offset instructions, process parameter switching instructions, and work instruction update instructions, and synchronously send the instructions to the corresponding execution terminals.

[0011] In a preferred embodiment, when the real-time status identification module generates a real-time status tensor of the factory based on real-time equipment operating parameters, real-time material inventory levels, and order delivery priorities, it is specifically used for:

[0012] Real-time operating parameters of the equipment, including load, temperature and vibration, are collected from the equipment controllers in the factory, and the operating time-series vector of the factory is generated by indexing the equipment number.

[0013] Obtain the types and available quantities of materials in the factory's warehouse locations, and combine this with the consumption rate of the factory's line-side warehouses to generate the real-time inventory level of the factory's materials, thereby obtaining the factory's inventory status vector.

[0014] Extract the delivery time and customer level of the outstanding orders in the factory, and sort them according to their urgency to generate order delivery priorities, so as to form the order weight vector of the factory;

[0015] Align the operating condition time-series vector, the inventory status vector, and the order weight vector along the same time axis, and reconstruct them into a three-dimensional tensor using equipment number, material code, and order number as dimensions to generate the real-time status tensor of the factory.

[0016] In a preferred embodiment, when the degradation generation module performs continuous time-domain analysis on the real-time operating parameters of the equipment in the real-time state tensor to obtain the characteristic degradation curves of key moving parts of the equipment in the factory, it is specifically used for:

[0017] Extract time series slices of spindle load rate and feed axis vibration amplitude from the real-time state tensor;

[0018] By removing instantaneous load spikes in the spindle load rate time series caused by sudden changes in workpiece material or uneven cutting allowance, a smooth load baseline characterizing the basic energy consumption trend is obtained.

[0019] The characteristic frequency components associated with the fundamental frequency and harmonics of the rotating component in the time series of the vibration amplitude of the feed shaft are separated to obtain the vibration energy spectral density sequence reflecting the fatigue degree of the component surface;

[0020] The smoothed load baseline and the vibration energy spectral density sequence are fused according to a unified time index;

[0021] Using time as the independent variable and the degree of degradation as the dependent variable, the fused data is fitted to obtain the characteristic degradation curves of the key moving parts of the equipment in the factory.

[0022] In a preferred embodiment, when the degradation generation module performs a process of fitting the fused data with time as the independent variable and degradation degree as the dependent variable to obtain the characteristic degradation curves of the key moving parts of the equipment in the factory, it is specifically used for:

[0023] Calculate the first derivative of the characteristic degradation curve at the current moment to obtain the instantaneous degradation rate of the key moving parts of the equipment in the factory;

[0024] The instantaneous degradation rate is compared with a preset typical rate range for the healthy stage. When the instantaneous degradation rate exceeds the upper limit of the typical rate range for the healthy stage, it is determined that the key moving parts of the equipment have entered the early degradation stage from the normal operation stage.

[0025] Record the time node at which the determination occurs, and mark the coordinate point corresponding to the time node on the feature degradation curve.

[0026] In a preferred embodiment, when the health degradation module performs a prediction based on the inflection point of the characteristic degradation curve to indicate that equipment in the factory is about to enter a sub-healthy state, it is specifically used for:

[0027] Perform a second-order difference operation on the feature degradation curve to identify candidate locations where the concavity and convexity of the feature degradation curve change;

[0028] From the candidate positions, extreme points whose curvature values ​​exceed a preset curvature threshold are selected, and the horizontal coordinate time value corresponding to the extreme point is marked as the starting time when the key moving parts of the equipment deviate from the baseline operating characteristics.

[0029] Obtain the degradation degree value corresponding to the feature degradation curve at the starting time, and compare the degradation degree value with the pre-calibrated sub-health state critical threshold;

[0030] When the degree of degradation reaches or exceeds the critical threshold of the sub-health state, the start time is directly determined as the warning timestamp;

[0031] When the degradation level value does not reach the critical threshold of the sub-health state, the degradation curve is extrapolated based on the average degradation rate after the starting time and extends to the intersection point of the critical threshold of the sub-health state. The intersection point time is determined as the precursor timestamp.

[0032] In a preferred embodiment, when the health degradation module performs the task of aggregating the precursor timestamps into a dynamic constraint matrix, it is specifically used for:

[0033] Establish a precursor timestamp sequence indexed by equipment identifier for the equipment in the factory, and write the precursor timestamps into the precursor timestamp sequence in chronological order;

[0034] Extract the start and end times of the processing periods that the equipment has scheduled for production in the future planning cycle from the real-time state tensor;

[0035] The time points in the precursor timestamp sequence are overlapped with the processing time period of the corresponding device. When the time point falls within the range of the processing time period, the processing time period is marked as a restricted time period.

[0036] Using the device identifier as the row index and the time unit within the planned period as the column index, construct the occupancy status table of the device, set the cell corresponding to the restricted period to the unavailable state, and set the remaining cells to the available state, and generate the dynamic constraint matrix of the device.

[0037] In a preferred embodiment, when the planning preset module identifies spatiotemporal resource conflict nodes in the factory during the planning execution cycle due to the overlap between the precursor timestamp and the predetermined process processing time, based on the order delivery priority in the dynamic constraint matrix and the real-time state tensor, it is specifically used for:

[0038] Traverse the cells marked as unavailable in the dynamic constraint matrix and extract the corresponding device identifiers and the start and end points of the restricted time periods;

[0039] Read the predetermined processes assigned to the device identifier within the current scheduling scheme from the real-time state tensor, and obtain the planned start time and planned completion time of the predetermined processes;

[0040] The intersection of the restricted period and the time interval formed by the planned start time to the planned completion time is calculated. When there is an overlap between the two, it is determined that there is a conflict between the predetermined process and the restricted period in terms of time and space resource occupation.

[0041] The conflicting process number, corresponding equipment identifier, and time boundary of the conflict overlap interval are combined into a conflict event record, and the conflict event record is identified as a spatiotemporal resource conflict node.

[0042] In a preferred embodiment, when the planning preset module identifies spatiotemporal resource conflict nodes in the factory during the planned execution cycle due to the overlap between the precursor timestamp and the predetermined process processing time, based on the order delivery priority in the dynamic constraint matrix and the real-time state tensor, it is further specifically used for:

[0043] Extract the order delivery priority associated with the process number of the device from the real-time state tensor;

[0044] According to the preset urgency mapping rules, the remaining days of delivery and customer level weight in the order delivery priority are converted into quantitative scores that represent the risk of order delay treatment.

[0045] Based on the quantified scores, the spatiotemporal resource conflict nodes within the same planning period are sorted in descending order to generate a conflict processing priority sequence for the spatiotemporal resource conflict nodes.

[0046] The conflict handling priority sequence is appended to the corresponding spatiotemporal resource conflict node attribute field, and the conflict handling priority sequence is used as the basis for the execution order of subsequent redistribution schemes to handle conflict nodes in turn.

[0047] In a preferred embodiment, when the redistribution module performs the task of selecting alternative equipment with the same processing function and a current load rate lower than a preset elastic threshold for the conflicting processes of the factory based on the spatiotemporal resource conflict nodes, and thus obtaining a redistribution scheme for the factory, it is specifically used for:

[0048] Extract the processing function type required by the conflicting process from the spatiotemporal resource conflict node, and retrieve all equipment with the processing function type from the equipment capability configuration table to obtain the candidate equipment list of the factory;

[0049] The current load rate of each device in the candidate device list is read from the real-time state tensor, and devices with load rates lower than a preset elastic threshold are filtered out to obtain the available device list of the factory.

[0050] Query the availability status of the available equipment list in the dynamic constraint matrix during the conflict period, remove the equipment marked as unavailable during the conflict period, and obtain the final list of available equipment for the factory;

[0051] Select the device with the lowest load rate from the final list of available devices as the replacement target. Combine the process identifier of the conflicting process, the original device identifier, the device identifier of the replacement target device, and the adjusted planned processing time into a redistribution record, and write the redistribution record into the redistribution scheme.

[0052] In a preferred embodiment, when the instruction execution module decomposes the redistribution scheme into material offset instructions, process parameter switching instructions, and work instruction update instructions, and synchronously sends the instructions to the corresponding execution terminals, it is specifically used for:

[0053] Extract the process identifier of the conflicting process, the equipment identifier of the substitute target equipment, and the adjusted planned processing time from the reassignment scheme to generate a material offset instruction for the material delivery unit;

[0054] Retrieve the set of process parameters required by the alternative target equipment to process the conflicting process from the equipment parameter library, and generate a process parameter switching instruction for the equipment controller;

[0055] The work drawings and operating specifications of the conflicting process are associated with the equipment identifier of the alternative target equipment to generate an update instruction for the operation terminal.

[0056] The material offset command, the process parameter switching command, and the work instruction update command are respectively routed to the corresponding material delivery unit controller, equipment controller, and operation terminal display screen, and the command reception confirmation signal returned by the execution end is received to complete the synchronous issuance closed loop of the command.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. This invention generates characteristic degradation curves by analyzing equipment operating parameters in the time domain, accurately predicts the timestamps of equipment sub-health signs, and constructs a dynamic constraint matrix. It also identifies spatiotemporal resource conflict nodes by combining order delivery priorities and sorts and processes them according to their priorities. This avoids the problem of time overlap between abnormal equipment status and process processing from the source, prevents unplanned production interruptions, and ensures that the scheduling plan is highly matched with the actual production status of the factory.

[0059] 2. This invention selects the optimal alternative equipment from multiple dimensions, including processing function, load rate, and dynamic availability, to form a scientific redistribution scheme. It can also break down the scheme into three types of instructions: materials, processes, and work instructions, and send them to the corresponding execution terminals simultaneously, forming a closed loop of instruction issuance. This enables flexible allocation of production resources and coordinated adjustment of each execution link, adapting to the flexible production needs of smart factories and ensuring efficient delivery of orders according to priority. Attached Figure Description

[0060] Figure 1 This is a system architecture diagram of a dynamic planning and scheduling system for a digital intelligent factory, provided in an embodiment of the present invention.

[0061] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments belong to some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0063] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “said” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0064] Depending on the context, the word "if" or "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0065] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0066] In practice, the server-side equipment deployed in a dynamic planning and scheduling system for a smart factory may consist of one or more devices. This system can be implemented as a business instance, a virtual machine, or hardware devices. For example, it can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, it can be understood as software deployed on a cloud node to provide a dynamic planning and scheduling system for each user. Alternatively, it can be implemented as a virtual machine deployed on one or more devices in a cloud node, with application software installed to manage each user. Or, it can be implemented as a server composed of numerous identical or different types of hardware devices, with one or more devices configured to provide a dynamic planning and scheduling system for each user.

[0067] In terms of implementation, the dynamic planning and scheduling system for a smart factory and the user terminal are mutually compatible. That is, if the dynamic planning and scheduling system for a smart factory is implemented as an application installed on a cloud service platform, then the user terminal is implemented as a client that establishes a communication connection with the application; or if the dynamic planning and scheduling system for a smart factory is implemented as a website, then the user terminal is implemented as a webpage; or if the dynamic planning and scheduling system for a smart factory is implemented as a cloud service platform, then the user terminal is implemented as a mini-program in an instant messaging application.

[0068] like Figure 1 The diagram shown is a system architecture diagram of a dynamic planning and scheduling system for a smart factory provided in an embodiment of the present invention.

[0069] The intelligent factory dynamic planning and scheduling system 100 described in this invention can be set up in a cloud server. In terms of implementation, it can be used as one or more service devices, or as an application installed in the cloud (e.g., a mobile service operator's server, server cluster, etc.), or it can be developed as a website. Depending on the functions implemented, the intelligent factory dynamic planning and scheduling system 100 may include a real-time status identification module 101, a degradation generation module 102, a health degradation module 103, a plan preset module 104, a reallocation module 105, and an instruction execution module 106. The modules described in this invention can also be called units, referring to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0070] In this embodiment of the invention, in a dynamic planning and scheduling system for a smart factory, each of the above-mentioned modules can be implemented independently and can call other modules. Here, "calling" can be understood as a module connecting to multiple modules of another type and providing corresponding services to those connected modules. In the dynamic planning and scheduling system for a smart factory provided by this embodiment of the invention, the applicability of the system architecture can be adjusted by adding modules and directly calling them without modifying the program code, achieving cluster-based horizontal expansion to quickly and flexibly expand the system. In practical applications, the above modules can be set in the same device or different devices, or they can be set in virtual devices, such as service instances in a cloud server.

[0071] The following describes, with reference to specific embodiments, each component and its specific workflow of a smart factory dynamic planning and scheduling system:

[0072] The real-time status recognition module 101 is used to generate a real-time status tensor of the factory based on the real-time operating parameters of the equipment, the real-time inventory level of the materials, and the order delivery priority in the factory.

[0073] In this embodiment of the invention, when the real-time status recognition module generates a real-time status tensor of the factory based on the real-time operating parameters of the equipment, the real-time inventory level of materials, and the order delivery priority, it is specifically used for:

[0074] Real-time operating parameters of the equipment, including load, temperature and vibration, are collected from the equipment controllers in the factory, and the operating time-series vector of the factory is generated by indexing the equipment number.

[0075] Obtain the types and available quantities of materials in the factory's warehouse locations, and combine this with the consumption rate of the factory's line-side warehouses to generate the real-time inventory level of the factory's materials, thereby obtaining the factory's inventory status vector.

[0076] Extract the delivery time and customer level of the outstanding orders in the factory, and sort them according to their urgency to generate order delivery priorities, so as to form the order weight vector of the factory;

[0077] Align the operating condition time-series vector, the inventory status vector, and the order weight vector along the same time axis, and reconstruct them into a three-dimensional tensor using equipment number, material code, and order number as dimensions to generate the real-time status tensor of the factory.

[0078] A stable and continuous data transmission connection is established with the equipment controller of each production device through the industrial communication link within the factory. Real-time operating condition parameters such as load status, temperature status, and vibration status generated by the equipment during actual operation are continuously read from the equipment controller at fixed time intervals. Each read operating condition parameter is bound one-to-one with the corresponding read time to ensure that each set of operating condition parameters has a unique corresponding time identifier. Then, using the unique equipment number of each device as the core index, all operating condition parameters of the same device bound to time identifiers are arranged in chronological order of occurrence. The time-series operating condition parameters corresponding to all devices are uniformly collected and organized, and finally, a complete operating condition time-series vector that can fully reflect the real-time operating status of all equipment in the factory is generated.

[0079] The system retrieves information on the specific types of materials stored in each warehouse location from the factory's warehouse management system. Simultaneously, it meticulously counts and tallies the actual available quantity of each material in stock to ensure the accuracy of material type and available quantity information. Then, by collecting data on the actual usage of various materials in the line-side warehouse during continuous production, it calculates the consumption rate of each material per unit of production time, thereby determining the corresponding line-side warehouse consumption rate for each material. Based on the available quantity of each material and its corresponding line-side warehouse consumption rate, it calculates the continuous state of the material's current inventory to meet production needs, forming the material's real-time inventory level. Finally, it integrates and organizes the real-time inventory levels of all materials in the factory according to their classification, ultimately generating an inventory status vector that comprehensively reflects the factory's material reserves and consumption.

[0080] The system extracts all outstanding orders that have not yet completed production, processing, and delivery processes from the factory's order management system. It then meticulously extracts and verifies the agreed delivery time and corresponding customer level information for each outstanding order. First, it categorizes orders by their delivery time, with orders closer to their due date having higher urgency. Next, it adjusts the order importance based on customer level, with higher-level customers having higher importance. Combining urgency and importance, all outstanding orders are sorted sequentially. The sorting results determine the delivery priority of each outstanding order. Finally, the delivery priorities of all orders are uniformly grouped according to their unique ownership, generating an order weight vector that accurately reflects the factory's order delivery priorities.

[0081] The time identifiers corresponding to all data within the operating condition time-series vector, inventory status vector, and order weight vector are uniformly calibrated and adjusted to ensure that each set of data within the three vectors is attached to the same factory production operation timeline, achieving complete alignment in the time dimension. This ensures that data at the same moment in different vectors can correspond to each other. Then, using the equipment number as the first integration dimension, the material code as the second integration dimension, and the order number as the third integration dimension, the time-aligned operating condition time-series vector, inventory status vector, and order weight vector are layer-by-layer spliced ​​and three-dimensionally combined according to the correspondence of the three dimensions. This orderly integration and reconstruction of vector information scattered across different data dimensions ultimately generates a real-time state tensor that can comprehensively and multidimensionally reflect the overall real-time operating status of the factory.

[0082] The degradation generation module 102 is used to perform continuous time-domain analysis on the real-time operating parameters of the equipment in the real-time state tensor to obtain the characteristic degradation curves of the key moving parts of the equipment in the factory.

[0083] In this embodiment of the invention, when the degradation generation module performs continuous time-domain analysis on the real-time operating parameters of the equipment in the real-time state tensor to obtain the characteristic degradation curves of the key moving parts of the equipment in the factory, it is specifically used for:

[0084] Extract time series slices of spindle load rate and feed axis vibration amplitude from the real-time state tensor;

[0085] By removing instantaneous load spikes in the spindle load rate time series caused by sudden changes in workpiece material or uneven cutting allowance, a smooth load baseline characterizing the basic energy consumption trend is obtained.

[0086] The characteristic frequency components associated with the fundamental frequency and harmonics of the rotating component in the time series of the vibration amplitude of the feed shaft are separated to obtain the vibration energy spectral density sequence reflecting the fatigue degree of the component surface;

[0087] The smoothed load baseline and the vibration energy spectral density sequence are fused according to a unified time index;

[0088] Using time as the independent variable and the degree of degradation as the dependent variable, the fused data is fitted to obtain the characteristic degradation curves of the key moving parts of the equipment in the factory.

[0089] When the degradation generation module performs a process of fitting the fused data with time as the independent variable and degradation degree as the dependent variable to obtain the characteristic degradation curves of key moving parts of the equipment in the factory, it is specifically used for:

[0090] Calculate the first derivative of the characteristic degradation curve at the current moment to obtain the instantaneous degradation rate of the key moving parts of the equipment in the factory;

[0091] The instantaneous degradation rate is compared with a preset typical rate range for the healthy stage. When the instantaneous degradation rate exceeds the upper limit of the typical rate range for the healthy stage, it is determined that the key moving parts of the equipment have entered the early degradation stage from the normal operation stage.

[0092] Record the time node at which the determination occurs, and mark the coordinate point corresponding to the time node on the feature degradation curve.

[0093] From the completed real-time state tensor containing multi-dimensional information on equipment, materials, and orders, the data modules associated with the key moving parts of the equipment are precisely located according to the data hierarchy and time dimension of the equipment. The load rate data of the spindle and the vibration amplitude data of the feed axis during operation are then selected separately. These two types of data are completely extracted in a continuous arrangement from time to time to form a time series slice of spindle load rate and feed axis vibration amplitude that perfectly match the time dimension of the real-time state tensor.

[0094] Each data point in the spindle load rate time series is compared with consecutive data points before and after it to determine the degree of deviation between the data point and the values ​​of the consecutive data points before and after it. When a data point shows a significant increase and rapid decrease in value within a short period of time, it is determined that the data is an instantaneous load spike caused by a sudden change in the workpiece material or uneven distribution of the cutting allowance. The spike data is replaced with the average value of the multiple stable data points before and after it. The process of eliminating and replacing all instantaneous load spikes in the entire spindle load rate time series is completed in sequence, and finally a smooth load baseline that can stably characterize the trend of changes in the basic energy consumption of the equipment is formed.

[0095] Continuous frequency feature analysis is performed on the time series of feed shaft vibration amplitude. The vibration amplitude data at each moment is broken down according to the frequency distribution law of the vibration signal. The vibration signal components directly related to the fundamental frequency generated by the rotation of the moving part and the harmonics after the fundamental frequency is multiplied are identified. These characteristic frequency components related to the fundamental frequency and harmonics are completely separated from the original feed shaft vibration amplitude time series. Then, the energy corresponding to each characteristic frequency component is sorted according to the energy magnitude and arranged in chronological order to form a vibration energy spectral density sequence that can directly reflect the degree of fatigue damage on the surface of the moving part.

[0096] The time nodes corresponding to each data point in the smoothed load baseline and the vibration energy spectral density sequence are extracted separately. Using the same time node as the docking basis, the data in the smoothed load baseline and the data in the vibration energy spectral density sequence are matched one by one. The two types of data under the same time node are bound and integrated to achieve complete fusion of the smoothed load baseline and the vibration energy spectral density sequence under a completely unified time index.

[0097] On the characteristic degradation curve representing the state of key moving parts of equipment, the curve data position corresponding to the current moment is determined. At the same time, the curve data points immediately in front of and immediately behind this position are selected. The numerical changes of the current data point and the two adjacent data points and the corresponding time span are compared. The rate of change of the curve at the current moment is determined by the magnitude of the change of the value over time. Finally, the instantaneous degradation rate of key moving parts of equipment in the factory is obtained.

[0098] The typical rate range of the health stage is retrieved in advance based on the normal operating standards of the equipment. The instantaneous degradation rate of the key moving parts of the equipment is calculated and directly compared with the upper limit value of the typical rate range of the health stage. When the instantaneous degradation rate is greater than the upper limit value of the typical rate range of the health stage, it is directly determined that the operating state of the key moving parts of the equipment has switched from the normal operating stage to the early degradation stage.

[0099] At the moment the determination of the operation stage of the key moving parts of the equipment is completed, the specific time node of the determination result is accurately recorded. The time node is precisely matched with the time dimension of the feature degradation curve, and the coordinate position on the feature degradation curve that perfectly matches the time node is found. The coordinate position is clearly and fixedly marked.

[0100] The health degradation module 103 is used to predict the warning timestamps of the equipment in the factory that are about to enter a sub-healthy state based on the inflection point of the characteristic degradation curve, and to collect the warning timestamps into a dynamic constraint matrix.

[0101] In this embodiment of the invention, when the health degradation module executes the prediction of the warning timestamp that the equipment in the factory is about to enter a sub-healthy state based on the inflection point of the characteristic degradation curve, it is specifically used for:

[0102] Perform a second-order difference operation on the feature degradation curve to identify candidate locations where the concavity and convexity of the feature degradation curve change;

[0103] From the candidate positions, extreme points whose curvature values ​​exceed a preset curvature threshold are selected, and the horizontal coordinate time value corresponding to the extreme point is marked as the starting time when the key moving parts of the equipment deviate from the baseline operating characteristics.

[0104] Obtain the degradation degree value corresponding to the feature degradation curve at the starting time, and compare the degradation degree value with the pre-calibrated sub-health state critical threshold;

[0105] When the degree of degradation reaches or exceeds the critical threshold of the sub-health state, the start time is directly determined as the warning timestamp;

[0106] When the degradation level value does not reach the critical threshold of the sub-health state, the degradation curve is extrapolated based on the average degradation rate after the starting time and extends to the intersection point of the critical threshold of the sub-health state. The intersection point time is determined as the precursor timestamp.

[0107] When the health degradation module performs the operation of aggregating the precursor timestamps into a dynamic constraint matrix, it is specifically used for:

[0108] Establish a precursor timestamp sequence indexed by equipment identifier for the equipment in the factory, and write the precursor timestamps into the precursor timestamp sequence in chronological order;

[0109] Extract the start and end times of the processing periods that the equipment has scheduled for production in the future planning cycle from the real-time state tensor;

[0110] The time points in the precursor timestamp sequence are overlapped with the processing time period of the corresponding device. When the time point falls within the range of the processing time period, the processing time period is marked as a restricted time period.

[0111] Using the device identifier as the row index and the time unit within the planned period as the column index, construct the occupancy status table of the device, set the cell corresponding to the restricted period to the unavailable state, and set the remaining cells to the available state, and generate the dynamic constraint matrix of the device.

[0112] All data points arranged sequentially in chronological order on the feature degradation curve are read in sequence. First, the numerical difference between each data point and its previous adjacent data point is calculated to obtain the first-order change difference. Then, the numerical difference between each pair of adjacent first-order change differences is calculated to obtain the second-order change difference. By monitoring the position where the positive and negative states of the second-order change difference change, the specific points where the concavity and convexity of the feature degradation curve change are determined. All points where the concavity and convexity change are summarized and organized to form candidate positions where the concavity and convexity of the curve change on the feature degradation curve.

[0113] For each candidate position, the curvature value is obtained by calculating the curvature tightness of the feature degradation curve at that position, taking into account the arrangement of multiple data points adjacent to that position. The curvature value corresponding to each candidate position is compared with the preset curvature threshold set in advance based on the normal operating state of the equipment. Candidate positions with curvature values ​​greater than the preset curvature threshold are selected as valid extreme points. The specific time values ​​corresponding to these valid extreme points on the horizontal axis of the feature degradation curve are extracted, and the time values ​​are marked as the starting time when the key moving parts of the equipment begin to deviate from the baseline operating characteristics.

[0114] The specific data point that corresponds exactly to the starting time is accurately located on the characteristic degradation curve. The degradation degree value of the key moving parts of the equipment corresponding to the data point is read. At the same time, the sub-health state critical threshold value, which has been pre-calibrated through long-term stable operation test of the equipment, is retrieved. The degradation degree value read is directly compared with the sub-health state critical threshold value.

[0115] When the value of the degradation level reaches or exceeds the critical threshold of the sub-healthy state, the starting time of the deviation of the previously marked key moving parts of the equipment from the baseline operating characteristics is directly determined as the precursor timestamp of the key moving parts of the equipment.

[0116] When the degradation level value does not reach the critical threshold of sub-health state, the degradation rate values ​​corresponding to all time periods on the characteristic degradation curve after the starting time are counted. After summing up all these degradation rate values, the average degradation rate value is calculated. The characteristic degradation curve is continuously extended outward according to the changing trend of the average degradation rate until the extended curve intersects with the critical threshold of sub-health state. The time corresponding to the intersection point is determined as the precursor timestamp of the key moving parts of the equipment.

[0117] For each independently operating piece of equipment within the factory, a unique data record sequence is created. The unique equipment identifier of each piece of equipment is used as the retrieval basis and index identifier for this data sequence. A precursor timestamp sequence indexed by the equipment identifier is constructed. All precursor timestamps corresponding to each piece of equipment are entered into the corresponding precursor timestamp sequence in chronological order of their occurrence.

[0118] From the real-time status tensor containing comprehensive information on equipment status, materials, and orders, the production scheduling data module of the corresponding equipment is located. All processing task periods that have been scheduled and are determined to be executed by the equipment in the future planning cycle are completely extracted. The specific start and end times of each processing task are determined, and the start and end times of the processing periods that have been scheduled by the equipment in the future planning cycle are obtained.

[0119] Each precursor time point stored in the precursor timestamp sequence is matched with the time range overlap of each scheduled processing period of the corresponding equipment. It is determined whether each precursor time point is between the start and end time of the corresponding processing period. When the precursor time point falls within the range of the processing period, the processing period is marked as a restricted period in which the equipment cannot perform processing tasks at full capacity.

[0120] Using the unique equipment identifier of each piece of equipment in the factory as the horizontal row index and each independent time unit within the future planning cycle as the vertical column index, an occupancy status table for the corresponding equipment is constructed. In the occupancy status table, the cells corresponding to the restricted time periods are set to the unavailable state, and the cells corresponding to the other normal processing time periods that are not marked as restricted time periods are set to the available state. Through this status setting method, a dynamic constraint matrix for the equipment that can guide the real-time adjustment of equipment production scheduling is finally generated.

[0121] The planning preset module 104 is used to identify spatiotemporal resource conflict nodes in the factory during the planning execution cycle, based on the order delivery priority in the dynamic constraint matrix and the real-time state tensor, due to the overlap between the precursor timestamp and the predetermined process processing time.

[0122] In this embodiment of the invention, when the planning preset module identifies spatiotemporal resource conflict nodes in the factory during the planned execution cycle due to the overlap between the precursor timestamp and the predetermined process processing time, based on the order delivery priority in the dynamic constraint matrix and the real-time state tensor, it is specifically used for:

[0123] Traverse the cells marked as unavailable in the dynamic constraint matrix and extract the corresponding device identifiers and the start and end points of the restricted time periods;

[0124] Read the predetermined processes assigned to the device identifier within the current scheduling scheme from the real-time state tensor, and obtain the planned start time and planned completion time of the predetermined processes;

[0125] The intersection of the restricted period and the time interval formed by the planned start time to the planned completion time is calculated. When there is an overlap between the two, it is determined that there is a conflict between the predetermined process and the restricted period in terms of time and space resource occupation.

[0126] The conflicting process number, corresponding equipment identifier, and time boundary of the conflict overlap interval are combined into a conflict event record, and the conflict event record is identified as a spatiotemporal resource conflict node.

[0127] The planning preset module, when executing the order delivery priority based on the dynamic constraint matrix and the real-time state tensor, and identifying spatiotemporal resource conflict nodes caused by the overlap of the precursor timestamp and the predetermined process processing time within the planned execution cycle, is also specifically used for:

[0128] Extract the order delivery priority associated with the process number of the device from the real-time state tensor;

[0129] According to the preset urgency mapping rules, the remaining days of delivery and customer level weight in the order delivery priority are converted into quantitative scores that represent the risk of order delay treatment.

[0130] Based on the quantified scores, the spatiotemporal resource conflict nodes within the same planning period are sorted in descending order to generate a conflict processing priority sequence for the spatiotemporal resource conflict nodes.

[0131] The conflict handling priority sequence is appended to the corresponding spatiotemporal resource conflict node attribute field, and the conflict handling priority sequence is used as the basis for the execution order of subsequent redistribution schemes to handle conflict nodes in turn.

[0132] The dynamic constraint matrix corresponding to each device is retrieved sequentially according to the device's arrangement order. The entire matrix is ​​traversed by proceeding row by row and then checking column by column. The operating status of each cell is identified and confirmed one by one. All cells marked as unavailable are accurately filtered out. The unique device identifier corresponding to each unavailable cell is extracted from the row index to which the cell belongs. The start time and end time of the restricted time period corresponding to each unavailable cell are determined from the column index to which the cell belongs. The device identifier, start time and end time of the restricted time period corresponding to each unavailable cell are bound one-to-one and uniformly collected.

[0133] Within the real-time status tensor that integrates production information from all dimensions, including equipment status, material inventory, and order scheduling, a data module specifically storing the factory's current production scheduling plan is located. Based on the previously extracted equipment identifiers, precise data matching is performed, and all predetermined processing steps specifically assigned to that equipment identifier in the currently executing scheduling plan are read. For each predetermined step assigned to that equipment, the pre-planned start time and planned completion time of that step are retrieved from the scheduling data to ensure that the planned time information of each predetermined step corresponds completely and accurately with the corresponding equipment identifier.

[0134] The extracted restricted time period is used as the boundary to form a complete time coverage range. The planned start time to the planned completion time of the established process is taken as the continuous production execution time range. The positional relationship between the front and back boundaries of the two time ranges is compared one by one to check whether the start time of the restricted time period falls within the process execution time range, whether the end time of the restricted time period falls within the process execution time range, or whether the process execution time falls completely within the restricted time period. When there is any form of overlap between the two time ranges, it is directly determined that there is a conflict between the established process and the corresponding restricted time period in terms of time and space resource occupation.

[0135] Extract the unique process number corresponding to the established process where there is a conflict of time and space resources, combine it with the equipment identifier involved in the conflict and the start and end time boundaries of the conflict overlap interval, integrate the three core contents according to a fixed information combination logic to form a complete conflict event record, add a unique status mark to the conflict event record, and officially identify the conflict event record as a time and space resource conflict node.

[0136] In the real-time state tensor, the order data section directly associated with the equipment and process number involved in the conflict is located. Through precise matching of equipment identifier and process number, the order information directly bound to the process number is filtered out. The order delivery priority directly associated with the process number is completely extracted from it, ensuring that there is no deviation between the extracted order delivery priority and the association between the equipment and process number.

[0137] The urgency mapping rules, which were formulated in advance based on the factory's production management requirements and order delivery standards, are retrieved. According to the corresponding conversion standards set in the rules, the remaining days of delivery in the order delivery priority are first converted into the corresponding urgency value, and then the customer level weight in the order delivery priority is converted into the corresponding importance value. The values ​​of the two dimensions are then integrated and converted to form a quantitative score that can directly reflect the magnitude of the risk of order delay delivery.

[0138] Collect the quantified scores of all spatiotemporal resource conflict nodes generated within the same planning period. Arrange the spatiotemporal resource conflict nodes with larger quantified scores at the beginning of the sequence and arrange the spatiotemporal resource conflict nodes with smaller quantified scores at the end of the sequence. Through this orderly arrangement from high to low, a complete sequence of conflict handling priorities for spatiotemporal resource conflict nodes is generated.

[0139] The generated conflict handling priority sequence is added one by one to the exclusive attribute field of each corresponding spatiotemporal resource conflict node, so that each spatiotemporal resource conflict node is fixedly bound to its own corresponding handling priority. When formulating a process and equipment resource reallocation plan, the order of this conflict handling priority sequence is directly used as the sole basis for the execution order of handling each spatiotemporal resource conflict node.

[0140] The redistribution module 105 is used to select alternative equipment with the same processing function and whose current load rate is lower than a preset elastic threshold for the conflicting processes of the factory based on the spatiotemporal resource conflict nodes, so as to obtain the redistribution scheme of the factory.

[0141] In this embodiment of the invention, when the redistribution module performs the task of selecting alternative equipment with the same processing function and whose current load rate is lower than a preset elastic threshold for the conflicting processes of the factory based on the spatiotemporal resource conflict nodes, and obtains the factory redistribution scheme, it is specifically used for:

[0142] Extract the processing function type required by the conflicting process from the spatiotemporal resource conflict node, and retrieve all equipment with the processing function type from the equipment capacity configuration table to obtain the candidate equipment list of the factory;

[0143] The current load rate of each device in the candidate device list is read from the real-time state tensor, and devices with load rates lower than a preset elastic threshold are filtered out to obtain the available device list of the factory.

[0144] Query the availability status of the available equipment list in the dynamic constraint matrix during the conflict period, remove the equipment marked as unavailable during the conflict period, and obtain the final list of available equipment for the factory;

[0145] Select the device with the lowest load rate from the final list of available devices as the replacement target. Combine the process identifier of the conflicting process, the original device identifier, the device identifier of the replacement target device, and the adjusted planned processing time into a redistribution record, and write the redistribution record into the redistribution scheme.

[0146] From the identified spatiotemporal resource conflict nodes, extract the processing function types that the conflicting processes corresponding to the nodes must meet during the production process. Retrieve the equipment capacity configuration table compiled in advance in the factory, search the equipment capacity configuration table one by one for the processing functions of each piece of equipment, filter out all the equipment that can meet the processing function type, and summarize and organize these equipment that meet the processing function requirements to obtain the candidate equipment list corresponding to the factory.

[0147] From the real-time status tensor that integrates information on equipment operating conditions, materials, and orders, the data module related to equipment operating load is located. The current actual operating load rate of each device in the candidate device list is read one by one. The current load rate of each device is compared with the pre-set elastic threshold, and all devices with a current load rate lower than the preset elastic threshold are filtered out. These devices that meet the load requirements are collected and aggregated to obtain the list of available devices for the factory.

[0148] Retrieve the dynamic constraint matrix corresponding to each device in the list of available devices. For each device, query its operational availability status during the conflict period corresponding to the spatiotemporal resource conflict node. Check the status identifier of the cell corresponding to the conflict period in the dynamic constraint matrix one by one. Remove devices marked as unavailable during the conflict period and keep devices marked as available during the conflict period. Organize and summarize these retained devices to obtain the final list of available devices for the factory.

[0149] In the final list of available equipment, the current load rate of each piece of equipment is compared one by one. The equipment with the lowest load rate is identified and designated as the replacement target. The process identifier specific to the conflicting process, the original equipment identifier originally assigned to the process, the equipment identifier corresponding to the replacement target equipment, and the planned processing time determined after adjusting the available time of the replacement target equipment are extracted. These four types of information are integrated into a redistribution record in a fixed combination. The generated redistribution record is then uniformly entered into the corresponding redistribution scheme.

[0150] The instruction execution module 106 is used to decompose the redistribution scheme into material offset instructions, process parameter switching instructions, and work instruction update instructions, and synchronously send the instructions to the corresponding execution terminals.

[0151] In this embodiment of the invention, when the instruction execution module decomposes the redistribution scheme into material offset instructions, process parameter switching instructions, and work instruction update instructions, and synchronously sends the instructions to the corresponding execution terminals, it is specifically used for:

[0152] Extract the process identifier of the conflicting process, the equipment identifier of the substitute target equipment, and the adjusted planned processing time from the reassignment scheme to generate a material offset instruction for the material delivery unit;

[0153] Retrieve the set of process parameters required by the alternative target equipment to process the conflicting process from the equipment parameter library, and generate a process parameter switching instruction for the equipment controller;

[0154] The work drawings and operating specifications of the conflicting process are associated with the equipment identifier of the alternative target equipment to generate an update instruction for the operation terminal.

[0155] The material offset command, the process parameter switching command, and the work instruction update command are respectively routed to the corresponding material delivery unit controller, equipment controller, and operation terminal display screen, and the command reception confirmation signal returned by the execution end is received to complete the synchronous issuance closed loop of the command.

[0156] From the established redistribution plan, the dedicated process identifiers corresponding to the conflicting processes with spatiotemporal resource conflicts are precisely extracted, as well as the dedicated equipment identifiers corresponding to the alternative target equipment that undertakes the processing tasks of the conflicting processes. The planned processing time period is determined after adjusting the availability status of the alternative target equipment. Based on the extracted process identifiers, equipment identifiers, and planned processing time periods, the target objects, corresponding processes, and execution times of material delivery are identified, and a material offset instruction specifically issued to the material delivery unit is generated.

[0157] The system retrieves the pre-built equipment parameter database within the factory, accurately locates the target equipment in the database according to its identification, and then filters and matches the content based on the processing requirements of the conflicting process. It then retrieves all the process parameters necessary for the target equipment to complete the conflicting process, summarizes and integrates these process parameters to form a complete and unified process parameter set, and generates a special process parameter switching instruction for the equipment controller based on this process parameter set.

[0158] Extract all the work drawings and standardized operation specifications required for the conflicting processes in the actual production and processing. Establish a fixed correspondence between these work drawings and operation specifications and the equipment identification of the replacement target equipment, so that the work instruction content is accurately bound to the target equipment, and generate a work instruction update instruction specifically issued to the operation terminal.

[0159] Through a dedicated communication transmission channel within the factory, the generated material offset instructions are routed to the corresponding material distribution unit controller, the process parameter switching instructions are routed to the corresponding equipment controller, and the work instruction update instructions are routed to the corresponding operation terminal display screen. After each execution terminal successfully receives the corresponding instruction, it receives instruction reception confirmation signals returned by the material distribution unit controller, equipment controller, and operation terminal display screen in real time, confirming that all instructions have been delivered to the corresponding execution terminal and received, thus completing the closed loop of synchronous issuance of all instructions.

[0160] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0161] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A dynamic planning and scheduling system for intelligent factories, characterized in that, The system includes a real-time status identification module, a degradation generation module, a health degradation module, a plan preset module, a reallocation module, and an instruction execution module, wherein: The real-time status recognition module is used to generate a real-time status tensor of the factory based on the real-time operating parameters of the equipment, the real-time inventory level of materials, and the order delivery priority in the factory. The degradation generation module is used to perform continuous time-domain analysis on the real-time operating parameters of the equipment in the real-time state tensor to obtain the characteristic degradation curves of the key moving parts of the equipment in the factory. The health degradation module is used to predict the warning timestamps of equipment in the factory entering a sub-healthy state based on the inflection points of the characteristic degradation curves, and to aggregate the warning timestamps into a dynamic constraint matrix, specifically for: Perform a second-order difference operation on the feature degradation curve to identify candidate locations where the concavity and convexity of the feature degradation curve change; From the candidate positions, extreme points whose curvature values ​​exceed a preset curvature threshold are selected, and the horizontal coordinate time value corresponding to the extreme point is marked as the starting time when the key moving parts of the equipment deviate from the baseline operating characteristics. Obtain the degradation degree value corresponding to the feature degradation curve at the starting time, and compare the degradation degree value with the pre-calibrated sub-health state critical threshold; When the degree of degradation reaches or exceeds the critical threshold of the sub-health state, the start time is directly determined as the warning timestamp; When the degradation degree value does not reach the critical threshold of the sub-health state, the degradation curve is extrapolated based on the average degradation rate after the start time and extended to the intersection point of the critical threshold of the sub-health state. The intersection point time is determined as the precursor timestamp. Establish a precursor timestamp sequence indexed by equipment identifier for the equipment in the factory, and write the precursor timestamps into the precursor timestamp sequence in chronological order; Extract the start and end times of the processing periods that the equipment has scheduled for production in the future planning cycle from the real-time state tensor; The time points in the precursor timestamp sequence are overlapped with the processing time period of the corresponding device. When the time point falls within the range of the processing time period, the processing time period is marked as a restricted time period. Using the device identifier as the row index and the time unit within the planning period as the column index, construct the occupancy status table of the device, set the cell corresponding to the restricted period to the unavailable state, and set the remaining cells to the available state, and generate the dynamic constraint matrix of the device; The planning preset module is used to identify spatiotemporal resource conflict nodes in the factory during the planning execution cycle, based on the order delivery priority in the dynamic constraint matrix and the real-time state tensor, caused by the overlap between the precursor timestamp and the predetermined process processing time. The reallocation module is used to select alternative equipment with the same processing function and whose current load rate is lower than a preset elastic threshold for the conflicting processes of the factory based on the spatiotemporal resource conflict nodes, so as to obtain the reallocation scheme of the factory. The instruction execution module is used to decompose the redistribution scheme into material offset instructions, process parameter switching instructions, and work instruction update instructions, and synchronously send the instructions to the corresponding execution terminals.

2. The intelligent factory dynamic planning and scheduling system as described in claim 1, characterized in that, When the real-time status recognition module generates a real-time status tensor for the factory based on real-time equipment operating parameters, real-time material inventory levels, and order delivery priorities, it is specifically used for: Real-time operating parameters of the equipment, including load, temperature and vibration, are collected from the equipment controllers in the factory, and the operating time-series vector of the factory is generated by indexing the equipment number. Obtain the types and available quantities of materials in the factory's warehouse locations, and combine this with the consumption rate of the factory's line-side warehouses to generate the real-time inventory level of the factory's materials, thereby obtaining the factory's inventory status vector. Extract the delivery time and customer level of the outstanding orders in the factory, and sort them according to their urgency to generate order delivery priorities, so as to form the order weight vector of the factory; Align the operating condition time-series vector, the inventory status vector, and the order weight vector along the same time axis, and reconstruct them into a three-dimensional tensor using equipment number, material code, and order number as dimensions to generate the real-time status tensor of the factory.

3. The intelligent factory dynamic planning and scheduling system as described in claim 1, characterized in that, When the degradation generation module performs continuous time-domain analysis on the real-time operating parameters of the equipment in the real-time state tensor to obtain the characteristic degradation curves of key moving parts of the equipment in the factory, it is specifically used for: Extract time series slices of spindle load rate and feed axis vibration amplitude from the real-time state tensor; By removing instantaneous load spikes in the spindle load rate time series caused by sudden changes in workpiece material or uneven cutting allowance, a smooth load baseline characterizing the basic energy consumption trend is obtained. The characteristic frequency components associated with the fundamental frequency and harmonics of the rotating component in the time series of the vibration amplitude of the feed shaft are separated to obtain the vibration energy spectral density sequence reflecting the fatigue degree of the component surface; The smoothed load baseline and the vibration energy spectral density sequence are fused according to a unified time index; Using time as the independent variable and the degree of degradation as the dependent variable, the fused data is fitted to obtain the characteristic degradation curves of the key moving parts of the equipment in the factory.

4. The intelligent factory dynamic planning and scheduling system as described in claim 3, characterized in that, When the degradation generation module performs a process of fitting the fused data with time as the independent variable and degradation degree as the dependent variable to obtain the characteristic degradation curves of key moving parts of the equipment in the factory, it is specifically used for: Calculate the first derivative of the feature degradation curve at the current moment to obtain the instantaneous degradation rate of the key moving parts of the equipment in the factory; The instantaneous degradation rate is compared with a preset typical rate range for the healthy stage. When the instantaneous degradation rate exceeds the upper limit of the typical rate range for the healthy stage, it is determined that the key moving parts of the equipment have entered the early degradation stage from the normal operation stage. Record the time node at which the determination occurs, and mark the coordinate point corresponding to the time node on the feature degradation curve.

5. The intelligent factory dynamic planning and scheduling system as described in claim 1, characterized in that, When the planning preset module executes the order delivery priority based on the dynamic constraint matrix and the real-time state tensor, and identifies spatiotemporal resource conflict nodes in the factory during the planned execution cycle caused by the overlap of the precursor timestamp and the predetermined process processing time, it is specifically used for: Traverse the cells marked as unavailable in the dynamic constraint matrix and extract the corresponding device identifiers and the start and end points of the restricted time periods; Read the predetermined processes assigned to the device identifier within the current scheduling scheme from the real-time state tensor, and obtain the planned start time and planned completion time of the predetermined processes; The intersection of the restricted period and the time interval formed by the planned start time to the planned completion time is calculated. When there is an overlap between the two, it is determined that there is a conflict between the predetermined process and the restricted period in terms of time and space resource occupation. The conflicting process number, corresponding equipment identifier, and time boundary of the conflict overlap interval are combined into a conflict event record, and the conflict event record is identified as a spatiotemporal resource conflict node.

6. The intelligent factory dynamic planning and scheduling system as described in claim 1, characterized in that, The planning preset module, when executing the order delivery priority based on the dynamic constraint matrix and the real-time state tensor, and identifying spatiotemporal resource conflict nodes caused by the overlap of the precursor timestamp and the predetermined process processing time within the planned execution cycle, is also specifically used for: Extract the order delivery priority associated with the process number of the device from the real-time state tensor; According to the preset urgency mapping rules, the remaining days of delivery and customer level weight in the order delivery priority are converted into quantitative scores that represent the risk of order delay treatment. Based on the quantified scores, the spatiotemporal resource conflict nodes within the same planning period are sorted in descending order to generate a conflict processing priority sequence for the spatiotemporal resource conflict nodes. The conflict handling priority sequence is appended to the corresponding spatiotemporal resource conflict node attribute field, and the conflict handling priority sequence is used as the basis for the execution order of subsequent redistribution schemes to handle conflict nodes in turn.

7. The intelligent factory dynamic planning and scheduling system as described in claim 1, characterized in that, When the reallocation module performs the task of selecting alternative equipment with the same processing function and whose current load rate is lower than a preset elastic threshold for the conflicting processes of the factory based on the spatiotemporal resource conflict nodes, and obtains the factory's reallocation scheme, it is specifically used for: Extract the processing function type required by the conflicting process from the spatiotemporal resource conflict node, and retrieve all equipment with the processing function type from the equipment capacity configuration table to obtain the candidate equipment list of the factory; The current load rate of each device in the candidate device list is read from the real-time state tensor, and devices with load rates lower than a preset elastic threshold are filtered out to obtain the available device list of the factory. Query the availability status of the available equipment list in the dynamic constraint matrix during the conflict period, remove the equipment marked as unavailable during the conflict period, and obtain the final list of available equipment for the factory; Select the device with the lowest load rate from the final list of available devices as the replacement target. Combine the process identifier of the conflicting process, the original device identifier, the device identifier of the replacement target device, and the adjusted planned processing time into a redistribution record, and write the redistribution record into the redistribution scheme.

8. The intelligent factory dynamic planning and scheduling system as described in claim 1, characterized in that, When the instruction execution module decomposes the redistribution scheme into material offset instructions, process parameter switching instructions, and work instruction update instructions, and synchronously sends these instructions to the corresponding execution terminals, it is specifically used for: Extract the process identifier of the conflicting process, the equipment identifier of the substitute target equipment, and the adjusted planned processing time from the reassignment scheme to generate a material offset instruction for the material delivery unit; Retrieve the set of process parameters required by the alternative target equipment to process the conflicting process from the equipment parameter library, and generate a process parameter switching instruction for the equipment controller; The work drawings and operating specifications of the conflicting process are associated with the equipment identifier of the alternative target equipment to generate an update instruction for the operation terminal. The material offset command, the process parameter switching command, and the work instruction update command are respectively routed to the corresponding material delivery unit controller, equipment controller, and operation terminal display screen, and the command reception confirmation signal returned by the execution end is received to complete the synchronous issuance closed loop of the command.

Images