Process management method and system for an automated production line

Abstract

The application discloses a process management method and system for an automatic production line, and relates to the technical field of automatic production, and comprises the following steps: projecting a discrete control state of a current original process into a continuous physical process initial state based on a continuous physical process state expression model; constructing a physical process attractor domain of a target process in the state expression model based on a process corresponding to a target intelligent manufacturing unit; generating a continuous transition process path based on the physical process initial state and the physical process attractor domain; segmenting and migrating control authority based on the sequence of a differential transition process; and re-projecting the continuous physical process state of a current work-in-process into a discrete control state corresponding to the target process as an initial control state of the target intelligent manufacturing unit. Through the mapping of the continuous physical process state and the discrete control state and the generation of the continuous transition path, the application realizes smooth control transition of online in-out column of the intelligent manufacturing unit, and effectively eliminates the fault surface between the physical state and the control system.

Description

Technical Field

[0001] This invention relates to the field of automated production technology, and more specifically, to a process control method and system for automated production lines. Background Technology

[0002] With the development of industrial automation and modular intelligent manufacturing, modern production lines are increasingly adopting modular intelligent manufacturing units to achieve flexible and scalable production organization. In such production lines, each intelligent manufacturing unit collaborates through standard interfaces to realize multi-process processing and online transfer of work-in-process. However, since the thermal state, stress state, clamping state, and process evolution state of work-in-process are all continuously changing physical processes, while the control system typically switches processes using discrete workstations and discrete control commands, a time continuity discontinuity occurs between the control system's perception of the process state and the actual physical state of the work-in-process when an intelligent manufacturing unit enters or leaves the line online.

[0003] This discontinuity can cause abrupt changes in the process of work-in-process at the moment of entry or exit, further leading to problems such as processing instability, quality fluctuations, and discretization of processing results. Existing technologies typically struggle to establish an effective mapping relationship between discrete control switching and continuous physical processes, thus failing to guarantee a smooth transition of the physical state of work-in-process and process consistency during dynamic switching of intelligent manufacturing units.

[0004] Therefore, how to achieve a high-precision mapping between the continuous physical process state and the discrete control state on an automated production line, and maintain a smooth transition of the process when intelligent manufacturing units enter or leave the line online, is a key technical problem that urgently needs to be solved in the current automated production process control.

[0005] To address the above problems, this invention proposes a solution. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a process control method and system for automated production lines. By establishing a mapping relationship between the discrete control state of work-in-process and the continuous physical process state, and predicting the control transition path based on the evolution of the physical process state, the continuous transition and smooth switching of control authority when intelligent manufacturing units enter or leave the production line online are realized, thereby solving the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A process control method for automated production lines includes the following steps: When a smart manufacturing unit triggers online entry or exit, based on a continuous physical process state expression model, the discrete control state of the current original process is projected as the initial state of the continuous physical process; based on the process corresponding to the target smart manufacturing unit, a physical process attraction domain for the target process is constructed in the continuous physical process state expression model; based on the initial state of the continuous physical process and the physical process attraction domain, a continuous transition process path is generated and decomposed into multiple differential transition processes; based on the sequence of the differential transition processes, the control authority between the original smart manufacturing unit and the target smart manufacturing unit is transferred in segments; after completing the segmented transfer, the initial state of the current work-in-process continuous physical process is reprojected as the discrete control state corresponding to the target process, which serves as the initial control state of the target smart manufacturing unit.

[0009] In a preferred embodiment, the continuous physical process state expression model is specifically constructed as follows: by collecting temperature, clamping force, and surface process state information at each process node of the work-in-process, and fusing the information to form corresponding temperature state components, force state components, and process state components; combining the continuous state components into a multidimensional continuous state vector to characterize the instantaneous physical process state of the work-in-process; using the multidimensional continuous state vector as the state variable, establishing differential equations for each continuous state component, and introducing the coupling relationship between thermal-mechanical-process state into the differential equations to form a continuous physical process state expression model.

[0010] In a preferred embodiment, projecting the discrete control state of the current original process into the starting state of the continuous physical process specifically involves: when an online entry or exit trigger event is detected in the intelligent manufacturing unit, the discrete control state parameters corresponding to the original process are read to form a set of discrete control state parameters for the current original process; based on the preset mapping relationship between the discrete control state parameters and the continuous physical process state expression model, the spindle speed and feed rate are converted into an initial temperature distribution, the clamping command is mapped to an initial clamping force value, and the process segment number is mapped to a process state component to generate initial values ​​for each continuous state component; the initial values ​​are substituted into the state differential equation of the continuous physical process state expression model to solve for the continuous physical process starting state vector.

[0011] In a preferred embodiment, constructing the physical process attraction domain of the target process in the continuous physical process state expression model specifically involves: collecting the temperature state component, clamping force state component, and process state component of the work-in-process under the target process to generate a continuous physical process state time series corresponding to the target process; determining the steady-state range of each continuous state component based on the continuous physical process state time series to form a steady-state physical state constraint set; applying a perturbation with the steady-state physical state constraint set as the center, and filtering out a set of continuous state points that can regress to the steady-state constraint set through state expression model evolution; and determining the set of continuous state points as the physical process attraction domain corresponding to the target process.

[0012] In a preferred embodiment, the step of selecting the set of continuous state points that can regress to the steady-state constraint set specifically involves: using the steady-state mean of each continuous state component in the steady-state physical state constraint set as the disturbance center, applying a disturbance of a preset amplitude along the direction of each continuous state component to generate multiple sets of disturbance initial state points; substituting each disturbance initial state point into the continuous physical process state expression model for evolution calculation to obtain the corresponding continuous state evolution trajectory; performing regression determination on each continuous state evolution trajectory, and determining that when the continuous state re-enters the range of the steady-state physical state constraint set within a preset evolution time, the corresponding disturbance initial point is determined to be a regressible state point; and collecting all disturbance continuous state initial points determined to be regressible state points to form a continuous state point set, which serves as the physical process attraction domain corresponding to the target process.

[0013] In a preferred embodiment, generating a continuous transition process path based on the initial state of the continuous physical process and the attraction domain of the physical process specifically involves: using the initial state of the physical process as the initial node and the attraction domain of the physical process as the target convergence interval, under the constraints of the continuous physical process state expression model, constructing a continuous state evolution search space pointing from the initial state of the physical process to the attraction domain of the physical process; within the search space, using the comprehensive weighted offset value of the temperature state change, the force state change, and the process state change as the state disturbance cost function, evaluating the cost of each candidate continuous state evolution path according to the principle of minimum state disturbance, and selecting the continuous state evolution path with the minimum comprehensive disturbance cost as the continuous transition process path.

[0014] In a preferred embodiment, the decomposition into multiple differential transition steps specifically involves: sampling the continuous physical process state along the continuous transition process path, and segmenting the continuous transition process path according to the continuous state change rate; when the continuous state change rate falls within the same threshold interval, the corresponding continuous state change interval is divided into the same continuous sub-process; each of the continuous sub-processes is determined as a differential transition step, and for each differential transition step, the corresponding temperature state correction amount, clamping force correction amount, and process state correction amount are extracted to form a set of independently executable control parameter correction amounts; according to the order of the differential transition steps in the continuous transition process path, the control parameter correction amounts are sorted to form a differential transition step execution sequence for subsequent control authority segmentation migration.

[0015] In a preferred embodiment, the segmented migration of control permissions between the original intelligent manufacturing unit and the target intelligent manufacturing unit based on the sequence of differential transition processes specifically involves: sequentially reading the control parameter correction amount corresponding to the current differential transition process according to the execution order of the differential transition processes, and using it as the control transition instruction for the current stage; within the execution cycle corresponding to the current differential transition process, based on the control transition instruction, adjusting the control output weight of the original intelligent manufacturing unit downward according to a preset decreasing ratio, while simultaneously adjusting the control output weight of the target intelligent manufacturing unit upward according to an increasing ratio corresponding to the decreasing ratio, forming a dual-unit collaborative control weight allocation for the current stage; when the continuous physical process state corresponding to the current differential transition process reaches the target state interval corresponding to that process, it is determined that the current differential transition process has been completed, and the process switches to the next differential transition process, repeating the above segmented migration process of control permissions until all differential transition processes have been completed.

[0016] In a preferred embodiment, the step of reprojecting the current continuous physical process start state of the work-in-process to the discrete control state corresponding to the target process specifically involves: when the target intelligent manufacturing unit is ready to take over control, acquiring the current continuous physical process state vector of the work-in-process; determining whether the continuous state falls within the physical process attraction domain corresponding to the target process; if the condition is met, triggering the reprojection of the continuous state to the discrete control state; calculating the discrete control setpoint based on the standard control parameter range of the target process and the continuous physical process state vector; combining the discrete control setpoint with the corresponding process segment number to form a complete discrete control state parameter set under the target process; and using the discrete control state parameter set as the initial control state of the target intelligent manufacturing unit, allowing it to directly enter a stable production control state.

[0017] The technical effects and advantages of the process control method and system for automated production lines of this invention are as follows:

[0018] This invention establishes a mapping relationship between the discrete control state of work-in-process and the continuous physical process state, and generates a continuous transition path based on the evolution prediction of the physical process state, thereby realizing a smooth control transition during the online entry or exit process of intelligent manufacturing units. By gradually migrating control authority through segmented differential transition processes, and reprojecting the continuous state into the discrete control state of the target process after the transition is completed, this invention can effectively eliminate the gap between the physical state of work-in-process and the cognition of the control system, reduce process mutations, improve processing stability and product quality consistency, and achieve continuous and efficient operation of automated production lines without additional downtime or intervention. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the process control method for automated production lines according to the present invention.

[0020] Figure 2 This is a schematic diagram of the process control system for automated production lines according to the present invention. Detailed Implementation

[0021] 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 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1, Figure 1 The present invention provides a process control method for automated production lines, comprising the following steps:

[0023] S1, When the intelligent manufacturing unit triggers online entry or exit, based on the continuous physical process state expression model, the discrete control state of the current original process is projected into the continuous physical process start state;

[0024] The starting state of the continuous physical process refers to the instantaneous physical state starting point of the work-in-process, calculated by the continuous physical process state expression model at the moment when the intelligent manufacturing unit enters or leaves the line online. This is based on the discrete control state parameters of the original process and combined with the real-time temperature state, clamping force state, and surface process state of the work-in-process. It serves as the initial state boundary condition for the continuous transition calculation and control takeover of the subsequent target process.

[0025] In this embodiment, the continuous physical process state representation model is specifically constructed as follows:

[0026] At each process node, including processing, clamping, transfer, and waiting, infrared temperature measurement units, clamping force sensors, and online surface roughness detection devices are arranged on the spindle, tooling fixtures, and transfer mechanism to collect in-process temperature distribution signals, clamping contact force signals, and surface process status signals in real time. Based on the unified clock of the production line main controller, various signals are synchronized at the millisecond level. At the same time, according to the preset temperature-force-process dimension mapping benchmark, various signals are normalized and converted to obtain a unified set of process parameters that can be directly used in calculations at the same time scale.

[0027] From the unified process parameter set, based on the work-in-process structural parameters and process parameters, the temperature data at different sampling locations are weighted and fused to form a temperature state component, the contact force data at each clamping point are combined to form a force state component, and the surface roughness, residual stress, and processing texture parameters are combined with the current process segment number and encoded into a multi-dimensional vector through a fixed sequence to form a process state component.

[0028] The temperature state component, stress state component, and process state component are combined into a multidimensional continuous state vector according to a fixed state sorting rule, which is used to characterize the instantaneous physical process state of the work-in-process at the current moment.

[0029] Using the multidimensional continuous state vector as the state variable, a heat conduction differential equation for the temperature state component is established based on the thermal properties and structural dimensions of the work-in-process material. A force equilibrium differential equation for the force state component is established based on the stiffness matrix and force constraint relationship of the clamping mechanism. An evolution differential equation for the process state component is established based on the evolution law of the influence of processing texture parameters on surface quality. At the same time, a coupling term for the influence of temperature on material stiffness and processing deformation is introduced into each differential equation, thereby forming a continuous physical process state expression model that can solve the linkage evolution of the work-in-process thermal state, force state, and process state.

[0030] In each differential equation, the influence of temperature state on material stiffness and processing deformation is established through coupling matrix or coupling function to establish the interaction relationship between temperature, stress and process state, which is used to continuously solve the thermal state, stress and process state of the work-in-process.

[0031] It should be noted that, in this embodiment, the establishment of the differential equations and their coupling relationships corresponding to the temperature state component, the force state component, and the process state component is carried out according to the following steps:

[0032] Based on the material type, structural dimensions and processing environment of the work-in-process, the work-in-process is regarded as a physical object composed of multiple continuous or discrete spatial units. In each spatial unit, a corresponding temperature state variable is defined to characterize the instantaneous thermal state of the unit during the processing. Based on the thermal conductivity, specific heat and heat source input characteristics of the material, the basic conduction relationship of temperature change over time is determined, thereby forming a temperature state evolution relationship that describes the continuous evolution of temperature state over time.

[0033] When establishing the force state evolution relationship, the clamping mechanism and the work-in-process are regarded as an interacting force system. Based on the stiffness distribution, constraint form and force path of the clamping structure, the clamping force at each clamping point and the elastic deformation generated by the work-in-process are established to correspond to each other. On this basis, the balance adjustment relationship of the force state changing with time is determined, thus forming the force state evolution relationship used to describe the dynamic change of the force state.

[0034] When establishing the evolution relationship of process state components, based on the process law that the surface roughness, residual stress and processing texture gradually evolve with the changes in cutting heat, mechanical load and processing stage during the processing, the current process state is used as the historical state input, and combined with the influence of temperature state and stress state on the surface formation process, the update relationship of the process state gradually evolves over time is constructed to describe the continuous change process of process quality state.

[0035] After establishing the above evolutionary relationships, the influence of temperature state on material stiffness and processing deformation is introduced as a coupling factor. By introducing a material property correction term caused by temperature state changes into the stress state evolution relationship, the stress state changes can be dynamically adjusted with temperature changes. At the same time, an influence factor jointly affected by temperature state and stress state is introduced into the process state evolution relationship to characterize the influence of thermo-mechanical condition changes on the evolution rate and direction of surface process state.

[0036] The process of projecting the discrete control state of the current original process into the starting state of the continuous physical process specifically involves:

[0037] When an online entry or exit trigger event is detected in the intelligent manufacturing unit, the discrete control status parameters corresponding to the original process are first read, including spindle speed, feed rate, clamping command status and current process segment number, to form a set of discrete control status parameters for the current original process.

[0038] Based on the preset mapping rules between each control parameter in the discrete control state parameter set and the corresponding continuous state component in the continuous physical process state expression model, the spindle speed and feed rate are taken as the main energy input parameters in the cutting process. According to the common influence of the two on the material removal power per unit time, the corresponding heat input intensity is calculated, and the heat input intensity is used as the initial heat source condition in the heat conduction differential equation. In the heat conduction differential equation, the heat input intensity is combined with the material thermophysical parameters and structural dimension parameters of the work-in-process to perform transient solution, thereby obtaining the initial temperature distribution of the work-in-process at the triggering entry or exit time. The clamping command state is mapped to the initial clamping force value, and the process segment number is mapped to the corresponding process state component, thereby generating the initial assignment parameters of each continuous state component. The preset mapping rules from discrete control parameters to continuous state components can be pre-calibrated or configured according to the specific equipment type and processing technology, and those skilled in the art can directly implement it based on the above mapping logic.

[0039] The initial assignment parameters of each continuous state component are substituted into the state differential equation of the continuous physical process state expression model for transient solution to obtain the instantaneous state vector of the continuous physical process corresponding to the trigger entry or exit time, and the instantaneous state vector of the continuous physical process is determined as the starting state of the continuous physical process.

[0040] The specific formula for calculating the heat input intensity is as follows:

[0041]

[0042]

[0043]

[0044]

[0045] The initial clamping force is calculated using the following formula:

[0046]

[0047] In the formula, For heat input intensity, The thermal conversion coefficient, For cutting power, For cutting force, For cutting linear velocity, The diameter of the cutting tool. spindle speed, The material specific cutting force coefficient, For cutting depth, For feed rate, This represents the initial clamping force. This is the friction correction coefficient. Let be the command pressure of the i-th clamping actuator. This corresponds to the effective working area of ​​the piston.

[0048] It should be noted that the thermal conversion coefficient is obtained as follows: Under standard machining conditions, the machine tool performs stable cutting on the target material at the set spindle speed and feed parameters, while the spindle input power is acquired in real time through the machine tool power acquisition module; during the cutting process, the real-time temperature rise data of the workpiece is collected by a temperature sensor arranged on the surface of the workpiece, and the actual heat absorbed by the workpiece within the corresponding time is calculated based on the workpiece mass and specific heat capacity; the ratio of the actual heat absorbed by the workpiece to the spindle input power within the corresponding time is calculated to obtain the thermal conversion coefficient.

[0049] The material specific cutting force coefficient is obtained as follows: Select the target work-in-process material and perform trial cutting under multiple combinations of different cutting depths and feed rates; during the trial cutting process, collect cutting force data under each working condition by a cutting force sensor installed on the tool or spindle end; calculate the ratio of the cutting force data to the corresponding cutting cross-sectional area to obtain the cutting force coefficient corresponding to the unit cutting cross-sectional area, and determine this coefficient as the material specific cutting force coefficient.

[0050] The friction correction coefficient is obtained as follows: when the clamping mechanism is in a standard clamping state, the clamping actuator clamps the workpiece according to multiple different command pressures; during the clamping process, the actual contact force under each command pressure is collected by the pressure sensor arranged at the clamping contact position; the ratio of the actual contact force to the theoretical clamping force under the corresponding command pressure is calculated to obtain the correction ratio of the clamping force, and the correction ratio is determined as the friction correction coefficient.

[0051] S2, based on the corresponding process of the target intelligent manufacturing unit, construct the physical process attraction domain of the target process in the continuous physical process state expression model;

[0052] In this embodiment, the physical process attraction domain, under the corresponding process of the target intelligent manufacturing unit, is the range of continuous values ​​of the temperature state, clamping force state and surface process state of the work-in-process when the work-in-process is in a normal and stable processing state, which can naturally return and remain stable for a long time. This range is used to define the target convergence interval of the continuous transition process.

[0053] The construction of the physical process attraction domain of the target process in the continuous physical process state expression model based on the corresponding process of the target intelligent manufacturing unit is as follows:

[0054] When the target intelligent manufacturing unit is in a stable production state, the temperature state component, clamping force state component and process state component of the work-in-process under the target process are continuously collected in multiple processing cycles to obtain the continuous physical process state time series corresponding to the target process.

[0055] For the state time series, calculate the steady-state mean and allowable fluctuation range of temperature state, the steady-state distribution range and force balance deviation threshold of clamping force state, and the stable characteristic range of process state to form a set of steady-state physical state constraints corresponding to the target process.

[0056] The set of steady-state physical constraints is substituted into the continuous physical process state expression model as stable convergence boundary conditions for each continuous state component under the target process, which is used to limit the convergent range of the continuous state under normal processing conditions.

[0057] Centered on the set of steady-state physical constraints, positive and negative perturbations of preset amplitude are applied to the temperature state, stress state and process state respectively, and evolution calculations are performed through the continuous physical process state expression model to select all continuous state points that regress to the set of steady-state physical constraints within a preset time of the evolution process.

[0058] The set of multidimensional continuous state points that can automatically revert to the set of steady-state physical state constraints after being disturbed is determined as the physical process attraction domain corresponding to the target process.

[0059] It should be noted that the process of selecting all continuous state points that revert to the set of steady-state physical state constraints within a preset time period during the evolution process specifically involves:

[0060] Using the steady-state mean of each continuous state component in the set of steady-state physical constraints as the disturbance center, the disturbance amount is superimposed in the positive and negative directions according to the preset disturbance step size on the corresponding values ​​of the temperature state component, the clamping force state component and the process state component, to generate multiple sets of disturbance continuous state initial points.

[0061] Substitute each of the initial points of the continuous disturbance state into the continuous physical process state expression model for time-domain evolution calculation to obtain the continuous state evolution trajectory corresponding to each disturbance state.

[0062] Regression determination is performed on the evolution trajectory of each continuous state. When the continuous state re-enters the range of the steady-state physical state constraint set within a preset evolution time, the corresponding disturbance initial point is determined to be a regressible state point.

[0063] All the disturbance continuous state initial points that are determined to be regressible state points are gathered to form a continuous state point set, and the continuous state point set is determined as the physical process attraction domain corresponding to the target process.

[0064] S3, based on the initial state of the continuous physical process and the attraction domain of the physical process, generates a continuous transition process path and decomposes it into multiple differential transition steps;

[0065] In this embodiment, the generation of a continuous transition process path based on the initial state of the continuous physical process and the attraction domain of the physical process specifically includes:

[0066] Using the initial state of the physical process as the initial node of the continuous state evolution and the attraction domain of the physical process as the target convergence interval, under the constraints of the continuous physical process state expression model, a continuous state evolution search space is constructed from the initial state of the physical process to the attraction domain of the physical process.

[0067] Within the search space, the combined weighted offset value of temperature state change, stress state change, and process state change is used as the state disturbance cost function. According to the principle of minimum state disturbance, the cost of each candidate continuous state evolution path is evaluated, and the continuous state evolution path with the minimum comprehensive disturbance cost is selected as the continuous transition process path.

[0068] The cost evaluation of each candidate continuous state evolution path according to the principle of minimum state perturbation is specifically as follows:

[0069] For each candidate continuous state evolution path that points from the initial state of the physical process to the attraction domain of the physical process, the corresponding continuous state vector is read point by point along the path in chronological order.

[0070] The temperature state change, clamping force state change, and process state change between two adjacent continuous state vectors are calculated respectively, and the temperature state change, force state change, and process state change are normalized to obtain the standardized disturbance amplitude of each state component on the path.

[0071] According to the pre-set temperature disturbance weight, force disturbance weight, and process disturbance weight, the amplitude of each standardized disturbance is weighted and summed to obtain the comprehensive state disturbance cost corresponding to the candidate continuous state evolution path.

[0072] The comprehensive state disturbance cost corresponding to each candidate continuous state evolution path is compared, and the candidate path with the smallest comprehensive state disturbance cost is selected as the optimal continuous transition process path that satisfies the minimum state disturbance principle.

[0073] The decomposition is divided into multiple differential transition processes, specifically:

[0074] The states of the continuous physical process are sampled sequentially along the continuous transition process path, and the rate of change of the state between the continuous state vectors corresponding to two adjacent sampling times is calculated.

[0075] The continuous transition process path is segmented according to the state change rate and the preset rate segmentation threshold. When the continuous state change rate falls into the same threshold range, the corresponding continuous state change range is divided into the same continuous subprocess.

[0076] Each of the continuous sub-processes is defined as a differential transition process, and for each differential transition process, the corresponding temperature state correction, clamping force correction, and process state correction are extracted to form a set of independently executable control parameter corrections.

[0077] According to the order of the differential transition process in the continuous transition process path, the correction amounts of each group of control parameters are sorted to form a differential transition process execution sequence for subsequent control authority segmentation migration.

[0078] It should be noted that the step of defining each of the continuous sub-processes as a differential transition step, and extracting the corresponding temperature state correction, clamping force correction, and process state correction for each differential transition step, specifically involves:

[0079] At the start of the current differential transition process, the real-time measured values ​​of temperature, clamping force, and process status of the work-in-process under the original intelligent manufacturing unit are obtained, and the target temperature setting value, target clamping force setting value, and target process status setting value corresponding to the differential transition process under the target intelligent manufacturing unit are obtained simultaneously.

[0080] Based on the deviation between the measured value and the corresponding target set value, the temperature deviation, the force deviation and the process state deviation are calculated respectively.

[0081] Based on the physical evolution constraints of the differential transition process, the rate of change and amplitude of each deviation are limited to obtain the temperature state correction, clamping force correction and process state correction that meet the requirements of continuous physical transition.

[0082] The temperature state correction, clamping force correction, and process state correction are combined to form the control parameter correction for the current differential transition process.

[0083] S4, based on the sequence of differential transition processes, performs segmented migration of control permissions between the original intelligent manufacturing unit and the target intelligent manufacturing unit;

[0084] In this embodiment, the segmented migration of control permissions between the original intelligent manufacturing unit and the target intelligent manufacturing unit based on the sequence of differential transition processes is specifically as follows:

[0085] According to the execution order of the differential transition process, the control parameter correction amount corresponding to the current differential transition process is read sequentially, and the control parameter correction amount is used as the control transition instruction for the current stage.

[0086] Within the execution cycle corresponding to the current differential transition process, based on the control transition instruction, the control output weight of the original intelligent manufacturing unit is reduced according to a preset decreasing ratio, while the control output weight of the target intelligent manufacturing unit is increased according to an increasing ratio corresponding to the decreasing ratio, thus forming the dual-unit collaborative control weight allocation in the current stage.

[0087] Under the weight allocation of the dual-unit collaborative control, corresponding control commands are simultaneously issued to the original intelligent manufacturing unit and the target intelligent manufacturing unit, so that the original intelligent manufacturing unit gradually withdraws from the dominant control of the work-in-process, while the target intelligent manufacturing unit gradually takes over the dominant control of the work-in-process.

[0088] When the current differential transition process is detected to have reached the target state range corresponding to the current process, it is determined that the current differential transition process has been completed and the process is switched to the next differential transition process. The above control authority segmentation migration process is repeated until all differential transition processes have been completed.

[0089] The weight allocation for the dual-unit collaborative control at the current stage is specifically as follows:

[0090] At the start of the current differential transition process, the initial control output weights of the original intelligent manufacturing unit are set based on historical data. The initial control output weights of the target intelligent manufacturing unit ;

[0091] Based on the execution time of the current differential transition process, and considering the control response bandwidth of the original intelligent manufacturing unit and the target intelligent manufacturing unit under the current operating conditions, the execution cycle is divided into [number] parts according to the principle of not exceeding the minimum control response time constant. Each sub-cycle is updated at equal intervals, and the weight adjustment step size for each sub-cycle is preset.

[0092] Within each control update sub-cycle, the control output weights of the original intelligent manufacturing unit and the target intelligent manufacturing unit are updated synchronously.

[0093] The control commands output by the original intelligent manufacturing unit and the control commands output by the target intelligent manufacturing unit are weighted and synthesized according to the updated weights to generate the collaborative control commands that act on the work-in-process in the current sub-cycle.

[0094] At the end of each sub-cycle, the temperature response, stress response and process status response of the work-in-process are monitored in real time. When the monitoring results meet the state convergence threshold of the current differential transition process, the next control update sub-cycle is entered until the collaborative control weight migration corresponding to the current differential transition process is completed.

[0095] The synchronization update is specifically performed as follows:

[0096]

[0097]

[0098]

[0099]

[0100] In the formula, , These are the updated control output weights for the original intelligent manufacturing unit and the target intelligent manufacturing unit, respectively. , The initial control output weights for the original intelligent manufacturing unit and the target intelligent manufacturing unit are... This is the current sub-cycle number. Adjust the step size for the weights corresponding to each sub-cycle. The number of update sub-cycles is controlled at equal intervals.

[0101] S5, after completing the segmented migration, reprojects the current continuous physical process start state of the work-in-process to the discrete control state corresponding to the target process, as the initial control state of the target intelligent manufacturing unit.

[0102] In this embodiment, the step of reprojecting the current continuous physical process start state of the work-in-process into the discrete control state corresponding to the target process specifically involves:

[0103] When the target intelligent manufacturing unit is ready to take over control, the current continuous physical process state vector of the work-in-process is collected, including temperature state components, clamping force state components and process state components.

[0104] Determine whether the continuous physical process state falls entirely within the physical process attraction domain corresponding to the target process. When it is determined that it is within the physical process attraction domain, trigger the reprojection operation from the continuous state to the discrete control state.

[0105] Based on the standard control parameter range of the target process, and with the continuous physical process state vector as input, the spindle speed setting value, clamping force setting value and feed rate setting value are calculated respectively.

[0106] The back-calculated spindle speed setting, clamping force setting, and feed rate setting are combined with the process segment number corresponding to the target process to form a complete set of discrete control state parameters for the target process.

[0107] The discrete control state parameter set is used as the initial control state of the target intelligent manufacturing unit, so that the target intelligent manufacturing unit can directly enter a stable production control state without undergoing an additional transition process.

[0108] Example 2, Figure 2 The present invention provides a process control system for automated production lines, comprising a continuous physical process initial state generation module, an attraction domain construction module, a differential transition process decomposition module, a control authority segmentation migration module, and a target process discrete control state initialization module.

[0109] Continuous physical process start state generation module: When the intelligent manufacturing unit triggers online entry or exit, it projects the discrete control state of the current original process into the continuous physical process start state based on the continuous physical process state expression model.

[0110] Attraction Domain Construction Module: Used to construct the physical process attraction domain of the target process in the continuous physical process state expression model based on the corresponding process of the target intelligent manufacturing unit;

[0111] Differential Transition Process Decomposition Module: Used to generate continuous transition process paths and decompose them into multiple differential transition processes based on the initial state of the continuous physical process and the attraction domain of the physical process;

[0112] Control permission segmented migration module: used to perform segmented migration of control permissions between the original intelligent manufacturing unit and the target intelligent manufacturing unit based on the sequence of differential transition processes;

[0113] Target process discrete control state initialization module: After the segmented migration is completed, the current continuous physical process start state of the work-in-process is reprojected into the discrete control state corresponding to the target process, which serves as the initial control state of the target intelligent manufacturing unit.

[0114] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0115] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0116] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0117] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0118] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0119] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A process control method for automated production lines, characterized in that, Includes the following steps: When the intelligent manufacturing unit triggers online entry or exit, it projects the discrete control state of the current original process into the starting state of the continuous physical process based on the continuous physical process state expression model. The continuous physical process state expression model is specifically constructed as follows: temperature, clamping force and surface process state information are collected at each process node of the work-in-process, and the information is fused to form corresponding temperature state components, force state components and process state components; the state components are combined into a multi-dimensional continuous state vector to characterize the instantaneous physical process state of the work-in-process. Using multidimensional continuous state vectors as state variables, differential equations for each continuous state component are established, and the coupling relationship between thermal-mechanical-process states is introduced into the differential equations to form a continuous physical process state expression model. Based on the corresponding process of the target intelligent manufacturing unit, the physical process attraction domain of the target process is constructed in the continuous physical process state expression model. Specifically, the temperature state component, clamping force state component and process state component of the work-in-process under the target process are collected to generate the continuous physical process state time series corresponding to the target process. The steady-state range of each continuous state component is determined based on the state time series of the continuous physical process, forming a set of steady-state physical state constraints. A perturbation is applied with the set of steady-state physical state constraints as the center, and the set of continuous state points that can regress to the set of steady-state constraints is selected through the evolution of the continuous physical process state expression model. The set of continuous state points is determined as the physical process attraction domain corresponding to the target process. Based on the initial state of the continuous physical process and the attraction domain of the physical process, a continuous transition process path is generated and decomposed into multiple differential transition steps. Specifically, the initial state of the physical process is used as the initial node and the attraction domain of the physical process is used as the target convergence interval. Under the constraints of the continuous physical process state expression model, a continuous state evolution search space from the initial state of the physical process to the attraction domain of the physical process is constructed. Within the search space, the weighted offset value of the temperature state change, the force state change, and the process state change is used as the state disturbance cost function. Following the principle of minimum state disturbance, the cost of each candidate continuous state evolution path is evaluated, and the continuous state evolution path with the minimum comprehensive disturbance cost is selected as the continuous transition process path. The decomposition into multiple differential transition steps specifically involves: sampling the continuous physical process state along the continuous transition process path and segmenting the continuous transition process path according to the continuous state change rate. When the continuous state change rate falls within the same threshold interval, the corresponding continuous state change interval is divided into the same continuous sub-process. Each continuous sub-process is determined as a differential transition step, and for each differential transition step, the corresponding temperature state correction, clamping force correction, and process state correction are extracted to form a set of independently executable control parameter corrections. The control parameter corrections are sorted according to the order of the differential transition steps in the continuous transition process path to form a differential transition step execution sequence for subsequent control authority segmentation migration. Based on the sequence of differential transition processes, the control permissions between the original intelligent manufacturing unit and the target intelligent manufacturing unit are transferred in segments. Specifically, according to the execution sequence of the differential transition processes, the control parameter correction amount corresponding to the current differential transition process is read sequentially and used as the control transition instruction for the current stage. Within the execution cycle corresponding to the current differential transition process, based on the control transition instruction, the control output weight of the original intelligent manufacturing unit is reduced according to a preset decreasing ratio, while the control output weight of the target intelligent manufacturing unit is increased according to an increasing ratio corresponding to the decreasing ratio, forming the dual-unit collaborative control weight allocation for the current stage. When the continuous physical process state corresponding to the current differential transition process reaches the target state interval corresponding to the process, it is determined that the current differential transition process has been completed, and the process switches to the next differential transition process. The above segmented transfer process of control permissions is repeated until all differential transition processes have been completed. After completing the segmented migration, the initial state of the current work-in-process continuous physical process is reprojected into the discrete control state corresponding to the target process, which serves as the initial control state of the target intelligent manufacturing unit.

2. The process control method for automated production lines according to claim 1, characterized in that, The process of projecting the discrete control state of the current original process into the starting state of the continuous physical process specifically involves: When an online entry or exit trigger event is detected in the intelligent manufacturing unit, the discrete control state parameters corresponding to the original process are read to form the current discrete control state parameter set of the original process. Based on the preset mapping relationship between the discrete control state parameters and the continuous physical process state expression model, the spindle speed and feed rate are converted into the initial temperature distribution, the clamping command is mapped to the initial clamping force value, the process segment number is mapped to the process state component, and the initial assignment of each continuous state component is generated. Substitute the initial parameters into the state differential equation of the continuous physical process state expression model and solve to obtain the initial state vector of the continuous physical process.

3. The process control method for automated production lines according to claim 1, characterized in that, The process of selecting the set of continuous state points that can regress to the set of steady-state constraints specifically involves: Using the steady-state mean of each continuous state component in the set of steady-state physical state constraints as the disturbance center, a disturbance of a preset amplitude is applied along the direction of each continuous state component to generate multiple sets of initial state points of disturbance. Each initial state point of the disturbance is substituted into the continuous physical process state expression model one by one for evolution calculation to obtain the corresponding continuous state evolution trajectory; Regression determination is performed on the evolution trajectory of each continuous state. When the continuous state re-enters the range of the steady-state physical state constraint set within a preset evolution time, the corresponding disturbance initial point is determined to be a regressible state point. All the initial points of the continuous perturbation states that are determined to be regressible are gathered to form a set of continuous state points, which serves as the physical process attraction domain corresponding to the target process.

4. The process control method for automated production lines according to claim 1, characterized in that, The reprojection of the current continuous physical process start state of the work-in-process to the discrete control state corresponding to the target process specifically includes: When the target intelligent manufacturing unit is ready to take over control, the current continuous physical process state vector of the work-in-process is collected. Determine whether the continuous state falls within the physical process attraction domain corresponding to the target process. If the condition is met, trigger the reprojection of the continuous state to the discrete control state. Based on the standard control parameter range of the target process, the discrete control setpoint is calculated by combining the state vector of the continuous physical process. The discrete control setpoints are combined with the corresponding process segment numbers to form a complete set of discrete control state parameters for the target process. The discrete control state parameter set is used as the initial control state of the target intelligent manufacturing unit, allowing it to directly enter a stable production control state.

5. A process control system for automated production lines, applied to the process control method for automated production lines as described in any one of claims 1-4, characterized in that, It includes a continuous physical process initial state generation module, an attraction domain construction module, a differential transition process decomposition module, a control authority segmentation migration module, and a target process discrete control state initialization module: Continuous physical process start state generation module: When the intelligent manufacturing unit triggers online entry or exit, it projects the discrete control state of the current original process into the continuous physical process start state based on the continuous physical process state expression model. Attraction Domain Construction Module: Used to construct the physical process attraction domain of the target process in the continuous physical process state expression model based on the corresponding process of the target intelligent manufacturing unit; Differential Transition Process Decomposition Module: Used to generate continuous transition process paths and decompose them into multiple differential transition processes based on the initial state of the continuous physical process and the attraction domain of the physical process; Control permission segmented migration module: used to perform segmented migration of control permissions between the original intelligent manufacturing unit and the target intelligent manufacturing unit based on the sequence of differential transition processes; Target process discrete control state initialization module: After the segmented migration is completed, the current continuous physical process start state of the work-in-process is reprojected into the discrete control state corresponding to the target process, which serves as the initial control state of the target intelligent manufacturing unit.

Images