Production plan scheduling collaborative optimization method and system based on MOM and multi-source data

Abstract

The invention relates to the technical field of intelligent manufacturing and industrial software, and discloses a production plan scheduling collaborative optimization method and system based on MOM and multi-source data, and the method comprises the steps: constructing an equipment initial time-varying capability constraint envelope; extracting processing task demand features and calculating a processing effect residual vector; superposing the residual vectors to an equipment state step by step based on a state evolution equation so as to update an envelope in real time, and executing inclusion detection of a demand feature and the envelope; when the detection result is that the detection result is not included, identifying a physical dimension causing envelope narrowing, and generating an active recovery virtual task insertion sequence to remodel an equipment state track until the inclusion requirement is met; and finally generating a machine instruction code to execute closed-loop control. According to the method, a bidirectional coupling mechanism of task execution and equipment physical state evolution is established, online active recovery of equipment capability is realized by using an active recovery strategy, and high precision and stability of heavy equipment manufacturing are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing and industrial software technology, specifically to a collaborative optimization method and system for production planning and scheduling based on MOM and multi-source data. Background Technology

[0002] Heavy equipment manufacturing, exemplified by coal mining equipment such as hydraulic supports and coal mining machines, represents a typical complex discrete manufacturing industry within the industrial manufacturing sector. It is characterized by complex product structures, high levels of customization, long production cycles, and the production of multiple varieties in small batches. In this type of manufacturing model, the manufacturing operations management platform, as the core hub connecting the enterprise resource planning layer and the workshop control layer, is responsible for translating production orders into specific execution instructions. Through scientific production planning and scheduling, and the rational allocation of manufacturing resources such as personnel, equipment, and materials, it plays a crucial role in ensuring the on-time delivery of large equipment, shortening manufacturing cycles, and reducing operating costs.

[0003] Existing production planning and scheduling technologies follow a hierarchical processing logic: they obtain the master production schedule and bill of materials from the ERP system, combine them with inventory data from the warehouse management system, and use an advanced planning and scheduling engine to generate shop floor work plans. Their working principle is largely based on static resource capacity constraint models, employing operations research algorithms or heuristic rules to sort and assign processes. Under ideal, stable conditions, they can calculate production scheduling schemes that satisfy basic process constraints and distribute them to the production floor for execution via Gantt charts, attempting to construct a deterministic, predefined production order.

[0004] Current scheduling models focus primarily on static order or process dimensions, lacking multi-layered linkage mechanisms at the project, order, and process levels. This makes it difficult for macro-level project milestones to effectively constrain micro-level process execution, and leads to a disconnect between upper and lower level plans. Furthermore, existing systems are often passive and lagging in data utilization, making it difficult to integrate multi-source heterogeneous data such as equipment IoT and material logistics in real time. Consequently, scheduling schemes cannot detect dynamic disturbances such as sudden equipment failures and material delays. Although the generated plans are theoretically optimal, their actual executability is poor, and once an anomaly occurs, a time-consuming global rescheduling is often required.

[0005] Therefore, the purpose of this invention is to provide a collaborative optimization method and system for production planning and scheduling based on MOM and multi-source data, so as to overcome the shortcomings of the prior art. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a collaborative optimization method and system for production planning and scheduling based on MOM and multi-source data. It solves the problems of fragmented planning hierarchy in production planning and scheduling systems, which makes it difficult to coordinate macro milestones and micro execution; inability to effectively integrate multi-source heterogeneous data to cope with dynamic production environments; lack of rapid adaptive decision-making response mechanisms in the face of frequent disturbances; and difficulty in closed-loop iterative optimization of human experience knowledge and algorithm models, resulting in poor executability of scheduling schemes and difficulty in continuously improving the level of intelligence.

[0007] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of the present invention provides a collaborative optimization method for production planning and scheduling based on MOM and multi-source data, comprising the following steps: S1: Collect real-time operating data of the manufacturing equipment, and combine it with a preset natural decay model to establish an initial time-varying capability constraint envelope of the manufacturing equipment in the future scheduling window. The initial time-varying capability constraint envelope defines the performance boundary set of the manufacturing equipment under no processing load interference. S2: Analyze the production orders to be scheduled, extract the processing task demand feature fingerprint representing the hard physical conditions of the processing task for each processing task, and calculate the processing effect residual vector representing the reverse effect of the processing task on the physical state of the manufacturing equipment. S3: In the process of generating candidate task sequences, the initial time-varying capability constraint envelope established in step S1 is used as the evolution benchmark. According to the state evolution equation, the processing effect residual vectors calculated in step S2 corresponding to the preceding tasks in the candidate task sequence are superimposed to the current state of the manufacturing equipment step by step. The time-varying capability constraint envelope of subsequent time steps is calculated and updated in real time. S4: Perform inclusion detection to determine whether the processing task requirement feature fingerprint obtained in step S2 for subsequent pending processing tasks falls completely inside the time-varying capability constraint envelope updated in step S3. S5: When the inclusion detection result is not included, identify the physical dimension that causes the updated time-varying capability constraint envelope to narrow, generate an active recovery virtual task that does not produce actual output and insert it into the candidate task sequence, and use the active recovery virtual task to reshape the state evolution trajectory of the manufacturing equipment until the reshaped time-varying capability constraint envelope can contain the processing task requirement feature fingerprint of the subsequent pending processing tasks. S6: The final optimized sequence containing the processing task and the active recovery virtual task is converted into machine instruction code and sent to the manufacturing equipment for closed-loop control.

[0008] Preferably, in step S1, the process of establishing the initial time-varying capability constraint envelope of the manufacturing equipment within the future scheduling window further includes drift prediction logic, specifically including: Calculate the time-varying gradient based on the continuous time series state vector of the manufacturing equipment, and construct the equipment state evolution trend equation at the current moment. The residual vector of the processing effect corresponding to the processing task to be executed is superimposed on the equipment state evolution trend equation to generate a synthetic trajectory curve. The geometric intersection of the synthetic trajectory curve and the dynamic safety warning boundary inside the initial time-varying capability constraint envelope is calculated, and the geometric intersection is marked as the drift critical time point. When the drift critical time point is earlier than the planned execution time of the subsequent processing task, a trigger signal is generated to activate the active recovery virtual task generation process in step S5.

[0009] Preferably, the specific method of step S2 includes: The process attributes and technical specifications of the workpieces in the production order to be scheduled are analyzed to identify the key geometric features and quality constraint parameters of the workpieces to be processed. The quality constraint parameters are quantized and mapped into a multi-dimensional vector to construct the feature fingerprint of the processing task requirements. The dimensions of the multi-dimensional vector include at least the geometric tolerance level, the surface roughness threshold, and the minimum dynamic stiffness requirement. The machining effect residual vector is calculated using a physical effect mapping algorithm. Based on the workpiece material properties and cutting parameters, the thermal deformation increment, stress accumulation, and tool wear increment during the machining process are derived, and the thermal deformation increment, stress accumulation, and tool wear increment are integrated into the machining effect residual vector.

[0010] Preferably, in step S3, the ranking strategy based on complementary effects during the generation of candidate task sequences specifically includes: Establish a task classification matrix and classify the processing tasks into heat accumulation tasks, heat dissipation tasks, stress loading tasks, and stress release tasks based on the key component directions of the residual vector of the processing effect. Real-time monitoring of the offset vector of the current physical state of the manufacturing equipment relative to the geometric center of the initial time-varying capability constraint envelope; The processing task that can cancel out the offset vector by the residual vector of the processing effect is preferentially selected as the execution task of the next time step, thereby maintaining the stability of the state trajectory integral within the initial time-varying capability constraint envelope.

[0011] Preferably, in step S5, the specific method for generating an active recovery virtual task that does not produce actual output and inserting it into the candidate task sequence includes: Based on the identified physical dimension that causes the narrowing, the corresponding equipment state failure mode is determined, the preset recovery strategy library is accessed and the matching active recovery virtual task template is called. The active recovery virtual task template limits the recovery action types, including idling cooling, stress relief operation or tool reset. The state difference between the current state of the manufacturing equipment after the update in step S3 and the target permissible state boundary required to include subsequent processing tasks is calculated. The necessary execution time of the active recovery virtual task is then calculated in reverse based on the state difference using a preset physical response function. The active recovery virtual task generated based on the active recovery virtual task template is inserted before the processing task node in the candidate task sequence whose inclusion detection result is not included, and the planned start time of the unincluded processing task node and all subsequent processing tasks is postponed according to the necessary execution time.

[0012] Preferably, in step S5, when determining the execution duration of the active recovery virtual task, a parameter correction step based on the phase synchronization principle is also performed, specifically including: Construct a discrete demand weight sequence for the pending production and processing tasks, and a continuous capability carrier function characterizing the performance fluctuation law of the manufacturing equipment; Calculate the time-domain matching degree integral between the discrete demand weight sequence and the continuous capability carrier function; When the result of the time-domain matching degree integration shows that the high-weight processing task falls in the trough interval of the continuous capability carrier function, the duration of the active recovery virtual task is adjusted, and the subsequent continuous capability carrier function is phase-shifted in the time domain until the time window of the high-weight processing task is aligned with the peak interval of the continuous capability carrier function.

[0013] Preferably, in step S3, the specific method for constructing the state evolution equation based on the dynamic hysteresis loop geometric model includes: Construct a state phase plane coordinate system with external excitation load as the abscissa and physical state deviation as the ordinate; The loading response path during the execution phase of the fitting processing task and the unloading recovery path during the intermittent phase are used to define the closed region enclosed by the loading response path and the unloading recovery path as a dynamic hysteresis loop. Calculate the geometric area of ​​the dynamic hysteresis loop to characterize the dissipated energy density of a single processing cycle; By introducing a preset damage constitutive mapping coefficient, the dissipation energy density calculation is transformed into a state drift increment that represents the irreversible degradation of equipment performance; Based on the state accumulation principle of discrete time steps, a recursive mathematical expression describing the evolution of the equipment state with the processing process is established. The recursive mathematical expression makes the equipment state of the next time step equal to the vector sum of the current equipment state and the state drift increment term, thereby completing the construction of the state evolution equation.

[0014] Preferably, in step S1, the specific method for using the fatigue damage index to perform long-term constraint feedback correction on the natural decay model includes: Extract the dissipated energy density calculated for each processing task in step S3, perform a weighted summation calculation to generate the fatigue damage index for the current scheduling window, and calculate the damage accumulation rate of the fatigue damage index over time. The damage accumulation rate is compared with the full life cycle maintenance planning curve of the manufacturing equipment; When the damage accumulation rate exceeds a preset safety threshold, the estimated shrinkage amount of the envelope for the next cycle is calculated based on the accumulated amount of the fatigue damage index, and the boundary of the initial time-varying capability constraint envelope is shrunk inward using the estimated shrinkage amount of the envelope.

[0015] Preferably, before step S1, an initial boundary state confirmation step is further included, which specifically includes: Collect residual physical state data of the manufacturing equipment after the end of the previous cycle; The manufacturing equipment is driven to execute standardized rapid calibration and detection commands, collect real-time response data, and compare it with theoretical benchmark data to generate a deviation matrix; By combining the residual physical state data and the deviation matrix, the initial capability state vector at the current moment is calculated, and the initial capability state vector is used as the time zero-point boundary condition for establishing the initial time-varying capability constraint envelope in step S1.

[0016] A second aspect of the present invention provides a collaborative optimization system for production planning and scheduling based on MOM and multi-source data, the system comprising: Edge-side multi-dimensional state sensing units are deployed on manufacturing equipment to collect vibration, load, and temperature field condition data. The capability envelope calculation and edge gateway subsystem is connected to the edge-side multi-dimensional state perception unit and is used to receive the operating condition data and process it to generate a time-varying capability constraint envelope. The collaborative optimization and decision server is connected to the capability envelope calculation and edge gateway subsystem and is used to perform logical operations such as task parsing, effect residual vector calculation, dynamic sequence inference, envelope conflict detection, and active recovery virtual task generation. The scheduling instruction and device control coordination interface is used to convert the final optimization sequence containing the active recovery virtual task generated by the collaborative optimization and decision server into a composite instruction and send it to the device controller.

[0017] This invention provides a method and system for collaborative optimization of production planning and scheduling based on MOM and multi-source data. It has the following beneficial effects: 1. This invention establishes an initial time-varying capability constraint envelope and introduces a superposition feedback mechanism of processing effect residual vectors. It quantifies in real time the reverse effect of the preceding task execution process on the physical state of the equipment and superimposes it into the state evolution equation. When generating candidate task sequences, it prioritizes the combination of tasks with complementary physical effects. This changes the traditional scheduling mode that only passively adapts to the static constraints of the equipment. It actively mitigates the unidirectional decline of equipment performance by utilizing the physical interaction between tasks, thereby maintaining the physical stability of the manufacturing system in long-cycle operation and improving the accuracy consistency in continuous processing scenarios.

[0018] 2. This invention triggers an active recovery virtual task generation logic when the inclusion detection result is negative. Based on the specific physical dimension that causes the envelope to narrow, it automatically matches and inserts functional actions that do not produce actual output but can reshape the equipment state. Based on the predictive feedforward intervention strategy, it can actively restore the equipment's capability state before the equipment performance drifts to the critical point, avoiding unplanned downtime or process degradation caused by passive alarms. While ensuring high-precision processing capabilities, it effectively extends the continuous availability time of the equipment.

[0019] 3. This invention constructs a scheduling correction method based on the phase synchronization principle, performs time-domain integral matching between the discrete demand weight sequence of processing tasks and the continuous capability carrier function that characterizes the fluctuation law of equipment performance, and uses virtual tasks as phase-shifting operators to fine-tune the time axis. This can force the execution window of high-weight critical tasks to be precisely aligned with the peak range of equipment physical performance, avoid the risk of high-value workpieces operating during the low performance period of equipment, and achieve optimal spatiotemporal coordination between the time dimension of production planning and the performance dimension of physical resources. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the hardware architecture of the proactive production planning and scheduling collaborative optimization system based on residual feedback of processing effect and time-varying capability envelope of the present invention. Figure 2 This is a schematic diagram of the overall scheduling process based on envelope evolution according to the present invention; Figure 3 This is a schematic diagram illustrating the core scheduling logic and optimization strategy of the present invention; Figure 4 This is a schematic diagram of the physical state modeling and evolution algorithm of the present invention.

[0021] Among them, 100 is the edge-side multi-dimensional state perception unit; 200 is the capability envelope calculation and edge gateway subsystem; 300 is the collaborative optimization and decision server; and 400 is the scheduling instruction and device control collaborative interface. Detailed Implementation

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. 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.

[0023] See attached document Figure 1 The hardware architecture mainly includes an edge-side multi-dimensional state sensing unit 100. This unit is directly installed or integrated into the heavy-duty manufacturing equipment on the workshop floor, serving as the data acquisition source for the system. The edge-side multi-dimensional state sensing unit 100 is equipped with various types of industrial-grade sensors, specifically including high-frequency vibration sensors, spindle power monitoring sensors, and multi-point temperature field sensors. These sensors are connected to the equipment's CNC system or physical components to acquire real-time operating condition data reflecting the equipment's physical performance.

[0024] The data collected by the edge-side multi-dimensional state sensing unit 100 covers the vibration spectrum, spindle load current changes, and thermal distribution of key structural components during the machining process. This multi-source heterogeneous data constitutes the basic physical input for the time-varying capability constraint envelope of the subsequent computing equipment, reflecting the equipment's accuracy maintenance capability and stability during continuous machining.

[0025] Capability envelope computing and edge gateway subsystem 200. The capability envelope computing and edge gateway subsystem 200 communicates with the edge-side multi-dimensional state perception unit 100 via an industrial bus, receiving and processing raw operating condition data. The capability envelope computing and edge gateway subsystem 200 integrates an edge computing processor, which runs a preset lightweight capability degradation model and signal processing algorithm.

[0026] The Capability Envelope Computing and Edge Gateway Subsystem 200 performs time-domain alignment, feature extraction, and data dimensionality reduction on the acquired high-frequency raw signals. The Capability Envelope Computing and Edge Gateway Subsystem 200 transforms complex physical signals into capability state vectors that characterize the device's performance boundaries at the current moment. The processed capability state vectors are transmitted to the upper-layer server via industrial Ethernet. This edge-side preprocessing mechanism reduces data transmission latency and the computational load on the cloud server.

[0027] Collaborative Optimization and Decision Server 300. The Collaborative Optimization and Decision Server 300 is the computing core of the manufacturing operations management platform, storing product process design data, production order data, and a historical processing effect database. The Collaborative Optimization and Decision Server 300 receives capability state vectors from the capability envelope calculation and edge gateway subsystem 200 as initial boundary conditions for scheduling calculations.

[0028] The Collaborative Optimization and Decision Server 300 runs a scheduling optimization algorithm engine, which, based on a pre-defined residual model of processing effects, extrapolates the dynamic impact of different task sequences on the equipment capacity envelope. During the calculation process, the Collaborative Optimization and Decision Server 300 treats the equipment's real-time capacity envelope as an endogenous variable that changes with task execution. By finding the optimal matching sequence between task characteristics and the capacity envelope, it generates production execution instructions that include specific processing task sequences and proactive recovery actions.

[0029] The scheduling instruction and equipment control coordination interface 400 establishes a bidirectional data interaction channel between the collaborative optimization and decision server 300 and the underlying manufacturing equipment control system. The scheduling instruction and equipment control coordination interface 400 is responsible for parsing the production execution instructions generated by the collaborative optimization and decision server 300 and distributing them to the controllers of specific machines.

[0030] The scheduling instruction and equipment control coordination interface 400 has a composite instruction issuance function, which can transmit both conventional product processing G-code and proactive restorative virtual task instructions for reshaping the equipment capability envelope. These virtual task instructions include idling cooling cycle instructions, stress relief operation instructions, or tool reset instructions, thereby enabling the scheduling system to proactively intervene in and control the physical equipment status.

[0031] See attached document Figure 2 The collaborative optimization and decision server 300 first performs multi-source data fusion and baseline envelope construction steps. The collaborative optimization and decision server 300 receives the real-time state vector of the device from the capability envelope calculation and edge gateway subsystem 200. This real-time state vector contains dimensionality-reduced spindle load, vibration spectrum, and thermal distribution characteristic data. The collaborative optimization and decision server 300 combines the device's physical characteristic parameters stored in its local database to establish a natural degradation model of the device under no-processing load conditions. Based on the natural degradation model, the collaborative optimization and decision server 300 calculates and generates the initial time-varying capability constraint envelope of the device within a preset future time window. This initial time-varying capability constraint envelope defines the set of maximum processing accuracy, maximum allowable load, and stability thresholds that the device can achieve at different time points.

[0032] The Collaborative Optimization and Decision Server 300 then performs task feature extraction and effect quantification steps. The Collaborative Optimization and Decision Server 300 parses the production order process documents to be scheduled, extracting dual feature data for each machining task. The first feature is the demand feature fingerprint, representing the hard physical conditions required to complete the task, including tolerance grade, surface roughness requirements, and cutting force requirements. The second feature is the influence demand feature fingerprint, representing the reverse effect of the task execution process on the physical state of the equipment. The Collaborative Optimization and Decision Server 300 uses a preset machining effect mapping algorithm to calculate the machining effect residual vector corresponding to each task. The machining effect residual vector quantifies the spindle thermal elongation increment, residual stress accumulation, and tool wear increment caused by task execution.

[0033] The collaborative optimization and decision server 300 then executes a dynamic sequence deduction step based on envelope deformation feedback. The collaborative optimization and decision server 300 generates candidate task sequences in the scheduling solution space and performs time-domain evolution simulations on these candidate task sequences. During the simulation, the collaborative optimization and decision server 300, according to the state evolution equation, superimposes the residual vector of the processing effect of the preceding task onto the current state of the device, and updates the time-varying capability constraint envelope for subsequent time steps in real time. The collaborative optimization and decision server 300 performs inclusion detection to determine whether the requirement feature fingerprint of subsequent tasks completely falls within the updated time-varying capability constraint envelope.

[0034] When the updated time-varying capability constraint envelope fails to contain the requirement feature fingerprint of subsequent tasks, the collaborative optimization and decision server 300 triggers an active recovery micro-scheduling step. The collaborative optimization and decision server 300 identifies the types of physical factors causing envelope narrowing, including thermal deformation deviations or micro-vibration accumulation. Based on the identification results, the collaborative optimization and decision server 300 generates an active recovery virtual task that does not produce actual output. This recovery virtual task corresponds to specific equipment action logic, including high-speed idling preheating, low-speed cooling cycle, or stress-relief operation.

[0035] The collaborative optimization and decision server 300 inserts the generated proactive recovery dummy tasks into specific time nodes of the candidate task sequence. The collaborative optimization and decision server 300 recalculates the device state evolution trajectory after the insertion of the dummy tasks until the reshaped time-varying capability constraint envelope can again contain the requirement feature fingerprints of subsequent high-precision tasks. Through this process, the collaborative optimization and decision server 300 locks in the optimal scheduling scheme that satisfies bidirectional constraint matching.

[0036] The collaborative optimization and decision server 300 transforms the determined final execution sequence, which includes processing tasks and proactive recovery virtual tasks, into machine instruction code. This machine instruction code is then sent to the underlying equipment controller via the scheduling instruction and equipment control collaboration interface 400, driving the manufacturing equipment to execute production operations according to the reconstructed envelope logic, thus achieving dynamic closed-loop control of the physical processing process and the scheduling plan.

[0037] See attached document Figure 4 The Capability Envelope Calculation and Edge Gateway Subsystem 200 constructs a multi-dimensional state space for the equipment's physical performance. The Capability Envelope Calculation and Edge Gateway Subsystem 200 selects key physical quantities that characterize the equipment's processing quality and operational stability as the coordinate dimensions of the state space. These dimensions specifically include the processing accuracy level, the maximum permissible cutting load, and the vibration stability threshold. The Capability Envelope Calculation and Edge Gateway Subsystem 200 defines a closed geometric region within the multi-dimensional state space as the effective processing capability range possessed by the equipment at a specific moment.

[0038] The Capability Envelope Calculation and Edge Gateway Subsystem 200 establishes a steady-state evolution model of the equipment based on natural physical laws. This model describes the natural drift of the equipment's physical state over time under no external processing load. The natural drift encompasses the thermal equilibrium curve of the equipment's spindle during no-load operation, the pressure fluctuation baseline of the hydraulic system, and the release trajectory of static structural stress. The Capability Envelope Calculation and Edge Gateway Subsystem 200 utilizes historical no-load time-series data collected by the edge-side multi-dimensional state sensing unit 100 to generate a time evolution function of the equipment's basic state through regression analysis.

[0039] The collaborative optimization and decision server 300 generates a static baseline envelope based on the time evolution function of the equipment's basic state. The server performs discretization calculations on the scheduling time axis to determine the upper and lower bound vectors of the equipment capability index for each time step. The server then connects the upper and lower bound vectors of a series of consecutive time steps in the time domain to form a time-varying static baseline envelope. This static baseline envelope characterizes the theoretical processing capacity boundary of the equipment under natural evolution conditions and serves as the calculation benchmark for subsequent superimposed processing effect residues.

[0040] The Collaborative Optimization and Decision Server 300 performs balance verification of the capability envelope. During the static baseline envelope construction process, the Collaborative Optimization and Decision Server 300 introduces physical constraint coupling relationships between device components to ensure that capability indicators in each dimension meet physical compatibility. When a certain dimension indicator approaches a critical threshold due to natural degradation, the Collaborative Optimization and Decision Server 300 automatically shrinks the boundary range of the associated dimension based on the physical constraint coupling relationship, maintaining the geometric continuity and physical rationality of the capability envelope in multidimensional space, and completing the initial balance construction of the time-varying capability envelope.

[0041] See attached document Figure 2 Before the start of the current scheduling cycle, the edge-side multi-dimensional state perception unit 100 collects residual physical state data of the equipment after the end of the previous production cycle. This residual physical state data includes residual heat distribution during equipment downtime, residual thermal deformation of the mechanical structure, and static pressure values ​​of the hydraulic system. The capability envelope calculation and edge gateway subsystem 200 receives this residual physical state data and retrieves the natural recovery characteristic curves of the equipment under different resting durations from its locally stored historical state database to assess the basic physical conditions of the equipment at the current moment.

[0042] The collaborative optimization and decision server 300 generates standardized rapid calibration probe commands. These commands are designed to verify the actual dynamic performance of the equipment by executing specific sequences of benchmark actions. The scheduling command and equipment control coordination interface 400 receives the rapid calibration probe commands and translates them into code executable by the equipment controller, driving the equipment to perform benchmark actions, including spindle idling, rapid traverse of each axis, and positioning accuracy detection.

[0043] The edge-side multi-dimensional state sensing unit 100 synchronously collects real-time response data of the equipment during the execution of reference actions, including vibration response spectrum, positioning error values ​​of each axis, and spindle temperature rise rate. The capability envelope calculation and edge gateway subsystem 200 compares and analyzes the collected real-time response data with preset standard theoretical benchmark data to calculate the deviation matrix between the equipment's current actual performance and theoretical performance. The deviation matrix reflects the performance drift caused by wear accumulation due to long-term operation and environmental factors.

[0044] The Capability Envelope Calculation and Edge Gateway Subsystem 200 combines residual physical state data and the deviation matrix, and uses a multi-dimensional state mapping algorithm to calculate and generate the initial capability state vector for the current moment. The initial capability state vector precisely quantifies the actual effective machining accuracy level, maximum available cutting load threshold, and dynamic stiffness coefficient of the equipment at the start of scheduling. The Capability Envelope Calculation and Edge Gateway Subsystem 200 transmits the initial capability state vector to the Collaborative Optimization and Decision Server 300.

[0045] The collaborative optimization and decision server 300 locks the received initial capability state vector as the time zero-point boundary condition for the time-varying capability constraint envelope evolution calculation. Based on the zero-point boundary condition, the collaborative optimization and decision server 300 initializes and corrects the envelope shape of subsequent time steps, ensuring that the capability constraint space on which the scheduling algorithm depends strictly corresponds to the actual current state of the physical equipment, thus eliminating the risk of scheduling and execution decoupling caused by idealized assumptions.

[0046] See attached document Figure 2 The Collaborative Optimization and Decision Server 300 performs process analysis on the production order. It retrieves the digital process files for the processes to be scheduled from the manufacturing operations management database to obtain physical process parameters. These parameters include workpiece material properties, depth of cut, feed rate, spindle speed, and continuous cutting duration. The Collaborative Optimization and Decision Server 300 defines these physical process parameters as external stimulus input data for the manufacturing equipment system.

[0047] The Collaborative Optimization and Decision Server 300 constructs a task requirement feature vector based on analytical physical process parameters. It identifies the minimum equipment capabilities required to complete the process and extracts key constraint indicators. These key constraint indicators include geometric tolerance requirements, surface roughness thresholds, minimum dynamic stiffness requirements, and peak instantaneous cutting force. The Collaborative Optimization and Decision Server 300 quantifies these indicators and encapsulates them into a multi-dimensional requirement vector. This multi-dimensional requirement vector defines the entry conditions for task execution; that is, the multi-dimensional requirement vector must be contained within the current capability envelope of the equipment in geometric space.

[0048] The collaborative optimization and decision server 300 synchronously constructs the impact feature vector of the task. Utilizing a pre-defined physical effect mapping algorithm, the collaborative optimization and decision server 300 calculates the physical impact of the process execution on the physical state of the equipment. The collaborative optimization and decision server 300 inputs the workpiece material removal rate and the cutting heat generation coefficient, and derives the spindle thermal elongation increment and bed thermal deformation distribution values ​​caused by the machining process through a thermo-mechanical coupling calculation model.

[0049] The Collaborative Optimization and Decision Server 300 quantifies and calculates the residual vector of machining effects. Combining historical response data from the edge-side multi-dimensional state sensing unit 100 for similar operating conditions, the Collaborative Optimization and Decision Server 300 assesses the residual vibration energy and accumulated structural stress caused by the process. The Collaborative Optimization and Decision Server 300 integrates the calculated thermal deformation increment, accumulated stress, and tool wear increment into a residual vector of machining effects.

[0050] The collaborative optimization and decision server 300 establishes a dynamic transfer function for task and equipment states. The transfer function describes the mathematical relationship between the residual vector of processing effects and the current time-varying capacity constraint envelope of the equipment, specifically manifested as the contraction, shift, or distortion of the envelope boundary. The collaborative optimization and decision server 300 marks each task to be scheduled as a reshaping operator of the equipment capacity envelope shape, serving as the computational basis for subsequent sequence optimization based on complementary effects.

[0051] See attached document Figure 3 The collaborative optimization and decision server 300 establishes a classification matrix of tasks to be scheduled based on physical effects. The collaborative optimization and decision server 300 traverses each task in the task pool to be scheduled, reading its corresponding residual vector of processing effects. Based on the positive and negative directions and magnitudes of the key components in the vector, the collaborative optimization and decision server 300 classifies tasks into different physical stimulus types. Specific physical stimulus types include heat accumulation tasks, heat dissipation tasks, stress loading tasks, and stress release tasks. The collaborative optimization and decision server 300 uses this to construct a complementary index of physical effects between tasks, identifying task pairs with opposite physical state evolution trends.

[0052] The collaborative optimization and decision server 300 evaluates the deviation of the device's current capability status in real time. It receives real-time status data from the capability envelope calculation and edge gateway subsystem 200 and calculates the offset vector of the current device's physical state point relative to the geometric center of the static baseline envelope. When the offset vector is detected to be monotonically increasing and approaching the envelope boundary in a certain dimension, the collaborative optimization and decision server 300 locks the dimension as the current scheduling's primary constraint dimension.

[0053] The Cooperative Optimization and Decision Server 300 performs sequence selection computation based on effect cancellation. The Cooperative Optimization and Decision Server 300 searches for candidate tasks in the opposite direction of the master constraint dimension. The Cooperative Optimization and Decision Server 300 prioritizes tasks whose residual processing effect vector can be vector subtracted from the current device offset vector. For example, when the device is in a high-elongation state, the Cooperative Optimization and Decision Server 300 searches for heat-dissipating tasks with low load, low speed, or long intermittent periods, using them as the system excitation input for the next time step.

[0054] The Cooperative Optimization and Decision Server 300 constructs a cumulative effect prediction model for sequence superposition. The Cooperative Optimization and Decision Server 300 treats candidate task sequences as a continuous burst of excitation pulses applied to the device system. Based on the principle of linear superposition or a nonlinear decay coupling model, the Cooperative Optimization and Decision Server 300 calculates the state trajectory integral throughout the entire sequence execution process. The Cooperative Optimization and Decision Server 300 selects combinations that ensure the state trajectory integral remains within the time-varying capability constraint envelope and that the device state returns to near the steady-state baseline at the end of the sequence.

[0055] The Collaborative Optimization and Decision Server 300 establishes a scheduling decision criterion based on dynamic equilibrium. Unlike traditional rules that only aim for the shortest completion time, the Collaborative Optimization and Decision Server 300 prioritizes the convergence of equipment physical states as the first priority in sequence ordering. It eliminates homogeneous task sequences that, although time-sensitive, lead to the cumulative effect of residual processing effects exceeding the capacity envelope, and forcibly inserts complementary tasks with physical compensation attributes to maintain the physical stability of the manufacturing system during long-cycle operation.

[0056] See attached document Figure 3 The collaborative optimization and decision server 300 initiates scheduling and deduction logic based on physical state evolution. The collaborative optimization and decision server 300 arranges candidate tasks sequentially on the virtual timeline and calls the processing effect residual vectors corresponding to each task. Based on a preset nonlinear superposition model, the collaborative optimization and decision server 300 establishes the state evolution equation: ; ; in, This represents the device state vector for the next time step. This represents the device state vector at the current time step. Represents the natural decay coefficient matrix. Indicates the current processing task The residual vector of processing effect, This represents the coupling coefficient matrix indicating the influence of processing. This represents a random disturbance term.

[0057] The Collaborative Optimization and Decision Server 300 uses the state evolution equation to progressively accumulate the thermal deformation increment, residual stress, and tool wear generated by the preceding tasks onto the current state reference of the equipment, and calculates the continuous evolution value of the equipment's physical state as the processing process progresses.

[0058] The Collaborative Optimization and Decision Server 300 calculates the dynamic deformation trajectory of the time-varying capability constraint envelope of the computing device. Based on the accumulated physical state parameters, the Collaborative Optimization and Decision Server 300 reconstructs the effective capability boundary of the device in real time for each future time step. Dynamic deformation specifically manifests as geometric contraction of the capability envelope in high-precision regions, threshold decrease in the load dimension, or center point drift in the spatial coordinate system. The Collaborative Optimization and Decision Server 300 quantifies this change in capability space caused by the reverse effect of processing tasks, determining the actual available processing range of the device at a specific point in time.

[0059] The Collaborative Optimization and Decision Server 300 performs geometric interference detection on the task requirements and capability boundaries. It extracts the requirement feature vectors for subsequent tasks, which are defined in a multi-dimensional vector space by the tolerance level, surface quality, and rigidity requirements of the task. The Collaborative Optimization and Decision Server 300 projects these requirement feature vectors into the time-varying capability constraint envelope coordinate system reconstructed at the same time step for spatial inclusion analysis.

[0060] The collaborative optimization and decision server 300 determines the conditions for envelope deformation conflicts. When the value of any dimension of the projected demand feature vector exceeds the instantaneous boundary range of the time-varying capability constraint envelope, the collaborative optimization and decision server 300 confirms that a parameter out-of-bounds conflict has occurred. The collaborative optimization and decision server 300 identifies this conflict as a state mismatch caused by the cumulative effect of continuous processing, resulting in the future performance of the equipment failing to meet the specific task process requirements.

[0061] The collaborative optimization and decision server 300 generates conflict feature descriptors. These descriptors record the limiting physical dimensions that lead to the conflict, specifically including dimensions of excessive thermal elongation or insufficient vibration stability. The collaborative optimization and decision server 300 uses these conflict feature descriptors as feedback signals to trigger subsequent proactive recovery micro-scheduling processes to correct the task sequence or insert restorative actions.

[0062] See attached document Figure 4 The collaborative optimization and decision server 300 establishes a time-varying gradient calculation model for the physical state parameters of the equipment. Based on the continuous time-series state vector uploaded by the capability envelope calculation and edge gateway subsystem 200, the collaborative optimization and decision server 300 uses the sliding window difference method to calculate the rate of change and acceleration of key capability indicators. Key capability indicators include spindle thermal elongation rate, vibration fundamental frequency drift acceleration, and tool wear rate. The collaborative optimization and decision server 300 uses the above calculation results to construct a trend equation for the evolution of the equipment state at the current moment. The trend equation characterizes the natural drift direction and velocity of the equipment under the current operating condition inertia.

[0063] The collaborative optimization and decision server 300 performs future state extrapolation calculations based on the planned sequence. It reads the residual vectors of processing effects for tasks to be executed in the current scheduling queue. The server then progressively superimposes these residual vectors onto the trajectory of the equipment state evolution trend equation according to predetermined execution time nodes. This generates a composite trajectory curve that includes the predicted equipment state at future discrete time points. The composite trajectory curve reflects the overall physical state trend resulting from the superposition of natural evolution laws and processing loads.

[0064] The collaborative optimization and decision server 300 sets dynamic safety margin thresholds and critical time point identification logic. Within the maximum allowable capacity boundary of the equipment, the collaborative optimization and decision server 300 sets a shrinkage ratio coefficient, defining a dynamic safety warning boundary. The collaborative optimization and decision server 300 calculates the geometric intersection of the synthetic trajectory curve and the dynamic safety warning boundary on the future time axis. The collaborative optimization and decision server 300 marks the time value corresponding to the geometric intersection point as the drift critical time point, which represents the specific moment when, without intervention, the equipment capacity envelope will narrow to the point where processing quality requirements cannot be met.

[0065] The collaborative optimization and decision server 300 constructs a triggering and judgment mechanism based on time window interference. The collaborative optimization and decision server 300 compares the calculated drift critical time point with the planned execution time window of subsequent high-precision tasks. When the drift critical time point falls within the execution period of a high-precision task, or is earlier than the start time of the high-precision task, the collaborative optimization and decision server 300 determines that there is a risk of timing capability failure, meaning that the physical state of the device during the execution of the high-precision task will slip out of the effective control domain.

[0066] The collaborative optimization and decision server 300 generates asynchronous micro-scheduling trigger signals. Unlike fixed-period rescheduling, the collaborative optimization and decision server 300 only activates the active recovery scheduling module immediately when a risk of capability failure is detected. The collaborative optimization and decision server 300 outputs event data packets containing the drift critical time point and the master control physical dimension that caused the drift. The signal forces the system to insert corresponding restorative dummy tasks during task gaps before the drift critical time point arrives, achieving prediction-based feedforward timing control.

[0067] See attached document Figure 2 The collaborative optimization and decision server 300 establishes a data alignment mechanism based on spatiotemporal synchronization. Before executing specific scheduling decisions, the collaborative optimization and decision server 300 first locks the predetermined execution window of the task to be evaluated on the planned time axis. Based on the time window, the collaborative optimization and decision server 300 sends a status request command to the capacity envelope calculation and edge gateway subsystem 200 to obtain the equipment predicted capacity status data for a specific time period. Simultaneously, the collaborative optimization and decision server 300 retrieves the static process parameters of the task to be evaluated from the process database, indexing and aligning the static task requirement data and the dynamic equipment status data on the same time base.

[0068] The Collaborative Optimization and Decision Server 300 performs vectorized reconstruction of the task requirement feature fingerprint. It reads the task's process document and extracts key parameters that impose rigid constraints on equipment performance, specifically including geometric accuracy tolerances, surface texture requirements, and upper limits of cutting force load. The Collaborative Optimization and Decision Server 300 transforms these discrete process parameters into a requirement feature vector in a multi-dimensional coordinate system. This requirement feature vector defines the minimum set of physical capabilities required for task execution, constituting the included objects in the decision-making process.

[0069] The collaborative optimization and decision server 300 synchronously acquires and parses the instantaneous capability envelope of the equipment. It receives real-time status data and trend prediction data uploaded by the capability envelope calculation and edge gateway subsystem 200. Based on this data, the collaborative optimization and decision server 300 constructs a time-varying capability constraint envelope entity within the current moment and a predetermined execution window. This time-varying capability constraint envelope entity consists of a series of inequalities defined in a multi-dimensional state space, establishing the maximum feasible machining domain for the equipment considering its current thermal state, tool wear state, and mechanical aging state.

[0070] The collaborative optimization and decision server 300 performs spatial projection matching between the demand vector and the capability envelope. The collaborative optimization and decision server 300 projects the reconstructed task demand feature vector onto the multidimensional state space containing the time-varying capability constraint envelope of the equipment. The collaborative optimization and decision server 300 calculates the Euclidean distance or generalized algebraic distance between the endpoint of the demand feature vector and the boundary plane of the time-varying capability constraint envelope. Through generalized algebraic distance calculation, the collaborative optimization and decision server 300 quantifies the physical margin between the task requirements and the current limit capability of the equipment.

[0071] The collaborative optimization and decision server 300 generates collaborative acquisition results that include safety margins. The server not only determines whether the demand feature vector lies within the envelope, but also calculates the deviation of the demand feature vector from the geometric center of the envelope and its approximation to the boundary, generating a safety margin index. The server compares the safety margin index with a preset process stability threshold. When the safety margin index is below the threshold, even if the task is theoretically feasible, the server still classifies the collaborative acquisition result as high-risk and outputs a corresponding risk indicator signal for subsequent scheduling algorithms to perform downgrade processing or trigger an active recovery mechanism.

[0072] See attached document Figure 3The Collaborative Optimization and Decision Server 300 first performs a source analysis of the physical constraints causing scheduling conflicts. When the drift prediction logic issues a capability failure warning, the Collaborative Optimization and Decision Server 300 analyzes the instantaneous state deviation vector to determine the specific failure mode. The types of failure modes include excessive spindle thermal elongation, substandard machine bed thermal balance, accumulated risk of tool micro-chipping, and excessive residual stress in the structure. Based on the failure mode, the Collaborative Optimization and Decision Server 300 indexes the corresponding proactive recovery virtual task template from the pre-set recovery strategy library.

[0073] The collaborative optimization and decision server 300 performs a quantitative mapping calculation of state deviation to the time dimension. It invokes a selected proactive recovery virtual task template, which contains a response function of the device's physical state to a specific recovery action. The collaborative optimization and decision server 300 substitutes the difference between the current device state and the target state required for subsequent tasks into the response function, and inversely calculates the execution time required to reach the target state. For example, the collaborative optimization and decision server 300 calculates the number of idling cooling cycles required to reduce the spindle temperature from its current high level to the precision machining allowable temperature, thereby determining the specific length of the virtual task on the time axis.

[0074] The Collaborative Optimization and Decision Server 300 constructs insertion windows on the timeline of the scheduling sequence. It locks the preceding nodes of high-risk tasks and embeds the calculated proactive recovery dummy tasks into these preceding node positions. Based on the execution duration of the dummy tasks, the Collaborative Optimization and Decision Server 300 shifts the start times of all subsequent pending production tasks backward, updating the global scheduling timetable. During the insertion process, the Collaborative Optimization and Decision Server 300 maintains logical constraints between tasks, ensuring that the intervention of dummy tasks does not disrupt the continuity of the production process.

[0075] The Collaborative Optimization and Decision Server 300 reconstructs the evolution trajectory of the device capability envelope after the insertion of virtual tasks. The Collaborative Optimization and Decision Server 300 treats proactively recovering virtual tasks as a special process with a "negative processing effect" or "state reset effect," substituting it into the state evolution equation for deduction. The Collaborative Optimization and Decision Server 300 simulates the changes in physical parameters of the device after executing virtual tasks, generating a repaired and corrected time-varying capability constraint envelope. The corrected envelope manifests as a capability recovery or error band narrowing in a specific physical dimension.

[0076] The collaborative optimization and decision server 300 performs a secondary verification and matching of the corrected envelope with the requirements of subsequent tasks. The collaborative optimization and decision server 300 then projects the requirement feature fingerprints of the subsequent high-precision tasks onto the new capability envelope after the active recovery operation. The collaborative optimization and decision server 300 confirms that the requirement feature fingerprints completely fall within the new envelope boundary, verifying that the active recovery dummy task successfully eliminated the potential parameter out-of-bounds risk.

[0077] The collaborative optimization and decision server 300 transforms verified proactive recovery virtual tasks into control instructions executable by the equipment. Based on the type of virtual task, the server generates corresponding CNC macro program code, producing M-code cooling cycle instructions or multi-axis linkage break-in instructions with specified speeds and durations. The server encapsulates the CNC macro program code instructions into the final production operation plan package and issues it through the scheduling instruction and equipment control collaboration interface 400, ensuring that the physical equipment strictly follows the scheduling logic to automatically adjust its status.

[0078] See attached document Figure 3 The Collaborative Optimization and Decision Server 300 constructs a demand weight sequence based on task priority. It iterates through the task queue awaiting scheduling, assigning a corresponding weight value to each task based on its process precision level and quality sensitivity. The Collaborative Optimization and Decision Server 300 assigns high-precision, critical tasks a high weight value and routine, ordinary tasks a low weight value, thereby generating a discrete reference weight sequence on the timeline. This reference weight sequence numerically represents the distribution of the scheduling system's demand intensity for high-performance equipment.

[0079] The collaborative optimization and decision server 300 synchronously extracts the fluctuation characteristics of the time-varying capability constraint envelope of the equipment and constructs a capability carrier function. The collaborative optimization and decision server 300 analyzes the historical and predicted data uploaded by the equipment capability envelope calculation and edge gateway subsystem 200 to identify the quasi-periodic fluctuation pattern exhibited by the equipment within the processing cycle. The collaborative optimization and decision server 300 fits the spindle thermal elongation period and vibration mode change period of the equipment into a continuous capability carrier function. The peaks of the capability carrier function correspond to the stable range of equipment performance, and the troughs correspond to the drift range of equipment performance.

[0080] The collaborative optimization and decision server 300 performs time-domain matching degree calculation of demand weights and capacity carriers. On the pre-simulation time axis, the collaborative optimization and decision server 300 performs point-to-point multiplication of the discrete reference weight sequence and the continuous capacity carrier function, and integrates the product result over time to calculate the phase alignment index. : ; in, This indicates the phase alignment index. Indicates the start time of the scheduling cycle. Indicates the end time of the scheduling cycle. This indicates the timeframe for the pending production and processing tasks. Discrete demand weight sequence, This represents the continuous capability carrier function after time-domain phase shifting. This indicates the time shift caused by introducing an active restorative dummy task.

[0081] The Collaborative Optimization and Decision Server 300 uses the time integral result as a phase alignment metric to evaluate scheduling quality. When the time window of a high-weight task falls within the trough of the capacity carrier function, the product result becomes smaller, indicating a phase mismatch, meaning that critical tasks are scheduled during periods of low equipment performance.

[0082] The Cooperative Optimization and Decision Server 300 introduces an active recovery virtual task as a phase-shifting operator. For the time window of phase mismatch, the Cooperative Optimization and Decision Server 300 calls the idle cooling or stress-relief virtual tasks generated in the preceding steps. Utilizing the non-processing duration and physical state reset characteristics of the virtual tasks, the Cooperative Optimization and Decision Server 300 performs time-domain shifting on the subsequent capability carrier function. By adjusting the insertion position and duration of the virtual tasks, the Cooperative Optimization and Decision Server 300 changes the phase angle of the equipment capability fluctuation relative to the task sequence, causing the subsequent capability peak to be delayed or advanced on the time axis.

[0083] The Cooperative Optimization and Decision Server 300 performs an iterative phase-locking search. It continuously adjusts the dummy task parameters and recalculates the integral matching degree between the corrected capacity carrier and the demand weight sequence. The Cooperative Optimization and Decision Server 300 uses a gradient ascent method to search for the maximum point of the phase alignment index. The Cooperative Optimization and Decision Server 300 determines that the scheduled phase synchronization is complete if and only if the time window of the high-weight task precisely covers the peak center of the corrected capacity carrier function.

[0084] The collaborative optimization and decision server 300 solidifies the final scheduling scheme after phase synchronization correction. The collaborative optimization and decision server 300 locks the task sequence and time nodes after phase synchronization correction. At this point, the execution window of the critical processing tasks strictly coincides with the optimal window of the equipment's physical performance. The collaborative optimization and decision server 300 outputs the final scheduling scheme, ensuring that the manufacturing process is carried out under the physical state with optimal performance stability. Processing accuracy is controlled through algorithm-level timing alignment.

[0085] See attached document Figure 4The capability envelope calculation and edge gateway subsystem 200 constructs a state phase plane coordinate system to describe the nonlinear response characteristics of the equipment. The capability envelope calculation and edge gateway subsystem 200 selects the external excitation load borne by the equipment as the horizontal axis variable, specifically characterized by the spindle load current value or cutting torque value. The capability envelope calculation and edge gateway subsystem 200 selects the physical state deviation generated by the equipment as the vertical axis variable, specifically characterized by the thermal deformation amount or vibration displacement amplitude of key structural components. The capability envelope calculation and edge gateway subsystem 200 maps the high-frequency discrete data points collected by the edge-side multi-dimensional state sensing unit 100 within a single machining task cycle to the state phase plane coordinate system.

[0086] Capacity envelope calculation and edge gateway subsystem 200 are used to generate loading response paths and unloading recovery paths. Capacity envelope calculation and edge gateway subsystem 200 identify data point sets corresponding to the processing task execution phase and use the least squares method to fit and generate loading response path curves. These curves describe the trajectory of the device's state changes during the continuous increase and maintenance phases of the load. Capacity envelope calculation and edge gateway subsystem 200 also identify data point sets corresponding to the intermittent phase after the processing task ends and fit and generate unloading recovery path curves. These curves describe the trajectory of the device's physical state returning to normal after the load is removed.

[0087] The Collaborative Optimization and Decision Server 300 defines the geometrically closed region of the dynamic hysteresis loop. Due to the thermal inertia and mechanical damping characteristics of the equipment's physical system, the loading response path curve and the unloading recovery path curve do not coincide on the phase plane. The Collaborative Optimization and Decision Server 300 defines the closed geometric region enclosed by connecting the two path curves end to end as the dynamic hysteresis loop. The Collaborative Optimization and Decision Server 300 calculates the maximum intercept of the dynamic hysteresis loop in the vertical direction, defining it as the maximum hysteresis width. The maximum hysteresis width quantifies the range of state uncertainty caused by different loading and unloading directions under the same load level.

[0088] The collaborative optimization and decision server 300 calculates the geometric area characteristics of the dynamic hysteresis loop. The server performs integration on the closed region to obtain the area value of the dynamic hysteresis loop. The area value characterizes the residual energy or accumulated error potential energy that the equipment system failed to dissipate in a single processing cycle. The server uses the area as a quantitative indicator to evaluate the strength of the equipment's state memory effect; a larger area indicates a stronger influence of previous historical loads on the current state of the equipment.

[0089] The collaborative optimization and decision server 300 establishes rules for the evolution of the hysteresis loop geometry with frequency. Based on the execution frequency of the task sequence, the collaborative optimization and decision server 300 adjusts the tilt angle and shape width characteristics of the hysteresis loop. When the task execution frequency increases, the collaborative optimization and decision server 300, according to a preset physical model, rotates and widens the origin of the hysteresis loop, reflecting the increased hysteresis in the device's state response under high-frequency excitation. Using the modified dynamic hysteresis loop geometry model, the collaborative optimization and decision server 300 predicts the actual starting state point when the next task intervenes. The actual starting state point is not located at the theoretical zero point, but at a specific position on the unloading recovery path curve.

[0090] See attached document Figure 4 The collaborative optimization and decision server 300 receives dynamic hysteresis loop geometric feature data transmitted by the capability envelope calculation and edge gateway subsystem 200. The dynamic hysteresis loop geometric feature data includes the area values ​​of the closed regions of the hysteresis loops generated for each discrete task in the scheduling sequence, as well as the maximum hysteresis width value. The collaborative optimization and decision server 300 defines the area of ​​the closed regions of the hysteresis loops as the dissipated energy density in a single machining cycle. Dissipated energy density characterizes the portion of energy that the equipment system fails to convert into cutting work during task execution and instead transforms into internal material damage or irreversible heat accumulation.

[0091] The Cooperative Optimization and Decision Server 300 performs cumulative calculations of the Fatigue Damage Index (FDI). Based on a linear damage accumulation rule or a nonlinear impact coupling model, the Cooperative Optimization and Decision Server 300 performs a weighted summation of the dissipated energy densities corresponding to all tasks within the scheduling sequence prediction time window. During the calculation process, the Cooperative Optimization and Decision Server 300 introduces a sensitivity coefficient determined by the current aging state of the equipment, transforming the physical energy values ​​into standardized FDI values. The FDI value characterizes the time integral of the cumulative damage potential energy caused by the scheduling scheme to the equipment.

[0092] The Collaborative Optimization and Decision Server 300 establishes a physical meaning mapping for the FDI index. The Collaborative Optimization and Decision Server 300 directly maps the calculated FDI index to the permanent decay rate of the time-varying capability constraint envelope. Unlike the temporary envelope contraction caused by thermal deformation, the FDI index characterizes the degree of irreversible consumption of equipment precision lifespan. A high FDI index means that although the current scheduling scheme has not triggered parameter out-of-bounds conflicts in the short term, it is rapidly depleting the equipment's guideway precision retention and spindle bearing lifespan.

[0093] The Collaborative Optimization and Decision Server 300 constructs a long-term constraint mechanism based on the FDI gradient. The Collaborative Optimization and Decision Server 300 calculates the slope of the FDI exponent's growth over time, i.e., the damage accumulation rate. The Collaborative Optimization and Decision Server 300 compares the damage accumulation rate with a preset equipment lifecycle maintenance planning curve. When the damage accumulation rate is detected to exceed the allowable threshold of the planning curve, the Collaborative Optimization and Decision Server 300 determines that the current scheduling scheme has excessive load intensity, posing a risk of excessively consuming the expected lifespan of the equipment.

[0094] The Cooperative Optimization and Decision Server 300 uses the FDI index to correct the static baseline envelope. Based on the cumulative amount of the FDI index, the Cooperative Optimization and Decision Server 300 calculates the estimated contraction of the static baseline envelope for the next production cycle. The Cooperative Optimization and Decision Server 300 feeds this long-term decline prediction back into the scheduling algorithm, forcing the algorithm to reduce the intensity of high-load tasks in the current sequence or insert dummy tasks of a deep maintenance nature, in order to smooth the growth slope of the FDI index and achieve a dynamic balance between short-term output efficiency and long-term equipment asset value.

[0095] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A collaborative optimization method for production planning and scheduling based on MOM and multi-source data, characterized in that, Includes the following steps: S1: Collect real-time operating data of the manufacturing equipment, and combine it with a preset natural decay model to establish an initial time-varying capability constraint envelope of the manufacturing equipment in the future scheduling window. The initial time-varying capability constraint envelope defines the performance boundary set of the manufacturing equipment under no processing load interference. S2: Analyze the production orders to be scheduled, extract the processing task demand feature fingerprint representing the hard physical conditions of the processing task for each processing task, and calculate the processing effect residual vector representing the reverse effect of the processing task on the physical state of the manufacturing equipment. S3: In the process of generating candidate task sequences, the initial time-varying capability constraint envelope established in step S1 is used as the evolution benchmark. According to the state evolution equation, the processing effect residual vectors calculated in step S2 corresponding to the preceding tasks in the candidate task sequence are superimposed to the current state of the manufacturing equipment step by step. The time-varying capability constraint envelope of subsequent time steps is calculated and updated in real time. S4: Perform inclusion detection to determine whether the processing task requirement feature fingerprint obtained in step S2 for subsequent pending processing tasks falls completely inside the time-varying capability constraint envelope updated in step S3. S5: When the inclusion detection result is not included, identify the physical dimension that causes the updated time-varying capability constraint envelope to narrow, generate an active recovery virtual task that does not produce actual output and insert it into the candidate task sequence, and use the active recovery virtual task to reshape the state evolution trajectory of the manufacturing equipment until the reshaped time-varying capability constraint envelope can contain the processing task requirement feature fingerprint of the subsequent pending processing tasks. S6: The final optimized sequence containing the processing task and the active recovery virtual task is converted into machine instruction code and sent to the manufacturing equipment for closed-loop control.

2. The production planning and scheduling collaborative optimization method based on MOM and multi-source data according to claim 1, characterized in that, In step S1, the process of establishing the initial time-varying capability constraint envelope of the manufacturing equipment within the future scheduling window also includes drift prediction logic, specifically including: Calculate the time-varying gradient based on the continuous time series state vector of the manufacturing equipment, and construct the equipment state evolution trend equation at the current moment. The residual vector of the processing effect corresponding to the processing task to be executed is superimposed on the equipment state evolution trend equation to generate a synthetic trajectory curve. The geometric intersection of the synthetic trajectory curve and the dynamic safety warning boundary inside the initial time-varying capability constraint envelope is calculated, and the geometric intersection is marked as the drift critical time point. When the drift critical time point is earlier than the planned execution time of the subsequent processing task, a trigger signal is generated to activate the active recovery virtual task generation process in step S5.

3. The production planning and scheduling collaborative optimization method based on MOM and multi-source data according to claim 1, characterized in that, The specific method of step S2 includes: The process attributes and technical specifications of the workpieces in the production order to be scheduled are analyzed to identify the key geometric features and quality constraint parameters of the workpieces to be processed. The quality constraint parameters are quantized and mapped into a multi-dimensional vector to construct the feature fingerprint of the processing task requirements. The dimensions of the multi-dimensional vector include at least the geometric tolerance level, the surface roughness threshold, and the minimum dynamic stiffness requirement. The machining effect residual vector is calculated using a physical effect mapping algorithm. Based on the workpiece material properties and cutting parameters, the thermal deformation increment, stress accumulation, and tool wear increment during the machining process are derived, and the thermal deformation increment, stress accumulation, and tool wear increment are integrated into the machining effect residual vector.

4. The production planning and scheduling collaborative optimization method based on MOM and multi-source data according to claim 1, characterized in that, In step S3, the ranking strategy based on complementary effects is executed when generating the candidate task sequence, specifically including: Establish a task classification matrix and classify the processing tasks into heat accumulation tasks, heat dissipation tasks, stress loading tasks, and stress release tasks based on the key component directions of the residual vector of the processing effect. Real-time monitoring of the offset vector of the current physical state of the manufacturing equipment relative to the geometric center of the initial time-varying capability constraint envelope; The processing task that can cancel out the offset vector by the residual vector of the processing effect is preferentially selected as the execution task of the next time step, thereby maintaining the stability of the state trajectory integral within the initial time-varying capability constraint envelope.

5. The production planning and scheduling collaborative optimization method based on MOM and multi-source data according to claim 1, characterized in that, In step S5, the specific method for generating an active recovery virtual task that does not produce actual output and inserting it into the candidate task sequence includes: Based on the identified physical dimension that causes the narrowing, the corresponding equipment state failure mode is determined, the preset recovery strategy library is accessed and the matching active recovery virtual task template is called. The active recovery virtual task template limits the recovery action types, including idling cooling, stress relief operation or tool reset. The state difference between the current state of the manufacturing equipment after the update in step S3 and the target permissible state boundary required to include subsequent processing tasks is calculated. The necessary execution time of the active recovery virtual task is then calculated in reverse based on the state difference using a preset physical response function. The active recovery virtual task generated based on the active recovery virtual task template is inserted before the processing task node in the candidate task sequence whose inclusion detection result is not included, and the planned start time of the unincluded processing task node and all subsequent processing tasks is postponed according to the necessary execution time.

6. The production planning and scheduling collaborative optimization method based on MOM and multi-source data according to claim 1, characterized in that, In step S5, when determining the execution duration of the active recovery virtual task, a parameter correction step based on the phase synchronization principle is also performed, specifically including: Construct a discrete demand weight sequence for the pending production and processing tasks, and a continuous capability carrier function characterizing the performance fluctuation law of the manufacturing equipment; Calculate the time-domain matching degree integral between the discrete demand weight sequence and the continuous capability carrier function; When the result of the time-domain matching degree integration shows that the high-weight processing task falls in the trough interval of the continuous capability carrier function, the duration of the active recovery virtual task is adjusted, and the subsequent continuous capability carrier function is phase-shifted in the time domain until the time window of the high-weight processing task is aligned with the peak interval of the continuous capability carrier function.

7. The production planning and scheduling collaborative optimization method based on MOM and multi-source data according to claim 1, characterized in that, In step S3, the specific method for constructing the state evolution equation based on the dynamic hysteresis loop geometric model includes: Construct a state phase plane coordinate system with external excitation load as the abscissa and physical state deviation as the ordinate; The loading response path during the execution phase of the fitting processing task and the unloading recovery path during the intermittent phase are used to define the closed region enclosed by the loading response path and the unloading recovery path as a dynamic hysteresis loop. Calculate the geometric area of ​​the dynamic hysteresis loop to characterize the dissipated energy density of a single processing cycle; By introducing a preset damage constitutive mapping coefficient, the dissipation energy density calculation is transformed into a state drift increment that represents the irreversible degradation of equipment performance. Based on the state accumulation principle of discrete time steps, a recursive mathematical expression describing the evolution of the equipment state with the processing process is established. The recursive mathematical expression makes the equipment state of the next time step equal to the vector sum of the current equipment state and the state drift increment term, thereby completing the construction of the state evolution equation.

8. The production planning and scheduling collaborative optimization method based on MOM and multi-source data according to claim 7, characterized in that, In step S1, the specific method for using the fatigue damage index to perform long-term constraint feedback correction on the natural decay model includes: Extract the dissipated energy density calculated for each processing task in step S3, perform weighted summation to generate the fatigue damage index for the current scheduling window, and calculate the damage accumulation rate of the fatigue damage index over time. The damage accumulation rate is compared with the full life cycle maintenance planning curve of the manufacturing equipment; When the damage accumulation rate exceeds a preset safety threshold, the estimated shrinkage amount of the envelope for the next cycle is calculated based on the accumulated amount of the fatigue damage index, and the boundary of the initial time-varying capability constraint envelope is shrunk inward using the estimated shrinkage amount of the envelope.

9. The production planning and scheduling collaborative optimization method based on MOM and multi-source data according to claim 1, characterized in that, Prior to step S1, an initial boundary state confirmation step is also included, which specifically includes: Collect residual physical state data of the manufacturing equipment after the end of the previous cycle; The manufacturing equipment is driven to execute standardized rapid calibration and detection commands, collect real-time response data, and compare it with theoretical benchmark data to generate a deviation matrix; By combining the residual physical state data and the deviation matrix, the initial capability state vector at the current moment is calculated, and the initial capability state vector is used as the time zero-point boundary condition for establishing the initial time-varying capability constraint envelope in step S1.

10. A production planning and scheduling collaborative optimization system based on MOM and multi-source data, characterized in that, The system is used to implement the production planning and scheduling collaborative optimization method based on MOM and multi-source data as described in any one of claims 1 to 9, the system comprising: An edge-side multi-dimensional state sensing unit (100) is deployed on the manufacturing equipment to collect vibration, load and temperature field condition data; The capability envelope calculation and edge gateway subsystem (200) is connected to the edge-side multi-dimensional state perception unit (100) and is used to receive the operating condition data and process it to generate a time-varying capability constraint envelope. The collaborative optimization and decision server (300) is connected to the capability envelope calculation and edge gateway subsystem (200) and is used to perform logical operations such as task parsing, effect residual vector calculation, dynamic sequence inference, envelope conflict detection and active recovery virtual task generation. The scheduling instruction and device control coordination interface (400) is used to convert the final optimization sequence containing the active recovery virtual task generated by the coordination optimization and decision server (300) into a composite instruction and send it to the device controller.

Images