An intelligent control system for processing parameters in copper busbar manufacturing.

Abstract

This invention relates to the field of metal forming and processing control technology, and discloses an intelligent control system for processing parameters in copper busbar manufacturing. The system includes a data mapping module that converts continuous thermal state data into discrete material slice data, monitors material transfer status, and outputs a comprehensive cooling intensity index and a material coordinate mapping table; a state decoupling module that decouples electrical feedback parameters, extracts equivalent forming load current or torque, and calculates the equivalent wear correlation index of the forming die; a feedforward control module that, combined with the above indices, mapping table, and global control gain parameters, outputs processing control commands with physical constraints; and a parameter optimization module that updates the gain parameters in a closed loop based on end-product inspection data. This invention solves the problems of spatial state mismatch and parameter coupling, achieves precise feedforward compensation of traction force, avoids traction force instability and jamming / breakage during operation, and effectively reduces the dimensional deviation rate of end products.

Description

Technical Field

[0001] This invention relates to the field of metal forming and processing control technology, specifically to an intelligent control system for processing parameters in copper busbar manufacturing. Background Technology

[0002] The manufacturing of copper busbars typically involves continuous processes such as continuous casting, continuous extrusion, forming, and annealing. In the forming process, the control of processing parameters directly affects the cross-sectional dimensions and quality of the final product. Traditional control systems often employ closed-loop feedback control, adjusting based on real-time feedback signals from equipment operation.

[0003] Existing control methods mostly collect time-based data directly when processing thermal state data, failing to adequately consider the physical displacement spans between different processes on the production line. Materials (such as copper rods or billets) must traverse a transport path from the pre-cooling process to the forming zone, causing a spatial mismatch between the real-time thermal parameters acquired by the control system and the current state of the material entering the forming mold. Under this mismatch, the system struggles to apply a matching traction force to compensate for the material's current cooling state.

[0004] When assessing forming load, existing technologies primarily rely on monitoring the total feedback current of the main drive system. However, in actual operation, the total current not only includes the forming deformation resistance but also incorporates the no-load component generated by the inherent friction of the mechanical transmission system and the inertial component during acceleration and deceleration. Because the inherent resistance of the mechanical system is not separated from the actual resistance caused by mold wear, parameter coupling occurs in the feedback data, making it difficult for the system to extract quantitative indicators that objectively reflect the mold wear state.

[0005] Conventional control often employs a reactive feedback mechanism, intervening only after dimensional deviations or traction fluctuations occur in the final product. When material resistance changes abruptly, this passive response mode cannot correct control commands in advance, leading to traction instability, material jamming, or breakage during operation. Furthermore, the control gain parameters of most processing equipment are fixed after initial settings, lacking a closed-loop iterative correction mechanism based on final quality inspection data. After long-term operation, the equipment cannot cope with the effects of mechanical aging and mold wear, ultimately resulting in out-of-tolerance product dimensions. Summary of the Invention

[0006] The technical problem addressed by this invention is that parameter control in existing copper busbar manufacturing processes typically relies on feedback from the quality of the final product or monitoring of a single electrical parameter. This lagging feedback control mode cannot track the spatial distribution of the thermal state of the copper busbar during multi-process transfer in real time, and it is difficult to effectively separate the frictional resistance and acceleration / deceleration inertial interference of the mechanical system from the actual wear state of the forming mold. These spatial state mismatches and parameter coupling problems prevent the system from accurately applying matching traction force compensation in the forming zone, leading to traction instability, material breakage, or out-of-tolerance dimensions of the final product during operation.

[0007] To address the above problems, the present invention provides the following technical solution: This invention provides an intelligent control system for processing parameters in copper busbar manufacturing, the system comprising: The data mapping module is used to receive on-site operating condition data, convert continuously collected thermal state data into discrete material slice data, monitor the material transfer status, and output a comprehensive cooling intensity index and a material coordinate mapping table. The state decoupling module is used to receive the field working condition data, process the electrical feedback parameters in the field working condition data, extract the equivalent forming load current or equivalent forming load torque, and calculate and output the equivalent wear correlation index of the forming mold. The feedforward control module is used to combine the comprehensive cooling intensity index, the material coordinate mapping table, the forming die equivalent wear correlation index, and the global control gain parameter to output processing control commands with physical constraints to the actuator. The parameter optimization module is used to remove and preprocess abnormal samples based on the terminal product quality inspection data corresponding to the processing control instructions after the batch production cycle is completed. After performing limit boundary verification and safety review, it outputs the updated global control gain parameters.

[0008] Furthermore, the data mapping module is used to receive the on-site operating condition data, convert the continuously collected thermal state data into discrete material slice data, and output the comprehensive cooling intensity index, specifically for: Based on the on-site working condition data, continuous copper rods or copper billets are divided into sets of virtual slices according to a fixed length. Using a normalization formula, the comprehensive cooling intensity index corresponding to each virtual slice is calculated based on the temperature and velocity data in the virtual slice set and the on-site working condition data.

[0009] Furthermore, the data mapping module is used to receive the on-site operating condition data, monitor the material transfer status, and, in conjunction with the already output comprehensive cooling intensity index, output the material coordinate mapping table, specifically for: Based on the physical inversion characteristics during the take-up and undo process, the virtual slice number sequence is reversed to form an initial logical mapping sequence. Based on the initial logical mapping sequence, combined with the meter counter feedback value in the field working data and the recorded shearing length, the target virtual slice number entering the forming area at the current moment is calculated using the slice index addressing formula. The comprehensive cooling intensity index of the corresponding slice is extracted and bound according to the target virtual slice number, and the material coordinate mapping table is updated and output.

[0010] Furthermore, the state decoupling module is used to receive the field operating condition data, process the electrical feedback parameters in the field operating condition data, extract the equivalent forming load current, calculate and output the equivalent wear correlation index of the forming mold, specifically for: The on-site working condition data is calculated using a current decoupling formula. The no-load current component that overcomes the frictional resistance of the mechanical system and the inertial compensation current component during acceleration and deceleration are deducted to extract the equivalent forming load current. The wear state evaluation formula is used to calculate the current equivalent wear correlation index of the forming mold based on the equivalent forming load current.

[0011] Furthermore, the state decoupling module is used to receive the on-site working condition data, calculate and output the equivalent wear correlation index of the forming mold, specifically for: The system monitors the angular acceleration, jamming discrete signal, and emergency stop status signal in the field operating data in real time to determine the current abnormal state of external interference. When the abnormal state of external interference is determined, the decoupling calculation is stopped, and the equivalent wear correlation index of the forming mold is maintained at the historical calculation value of the previous stable operating cycle.

[0012] Furthermore, the feedforward control module is used to combine the comprehensive cooling intensity index, the material coordinate mapping table, the equivalent wear correlation index of the forming die, and the global control gain parameter to output the processing control command with physical constraints to the actuator, specifically for: Based on the meter counter feedback value in the on-site working condition data, the physical slice currently entering the forming area is located in the material coordinate mapping table, and the comprehensive cooling intensity index corresponding to the physical slice is extracted; combined with the comprehensive cooling intensity index extracted by location, the current equivalent wear correlation index of the forming mold, and the first traction force compensation gain coefficient and the second traction force compensation gain coefficient in the global control gain parameters, the initial feedforward traction force command at the current sampling time is calculated using the feedforward traction force calculation formula.

[0013] Furthermore, after calculating the initial feedforward traction force command, the feedforward control module outputs the processing control command with physical constraints, specifically for: The initial feedforward traction force command is compared with the traction force safety upper limit threshold. When the initial feedforward traction force command is greater than or equal to the traction force safety upper limit threshold, the traction force protection mechanism is triggered to limit the actual output traction force command to the traction force safety upper limit threshold, generating a target feedforward traction force command. Based on the traction force over-limit difference between the initial feedforward traction force command and the traction force safety upper limit threshold, the cooperative deceleration compensation amount at the current sampling time is calculated using the cooperative deceleration calculation formula. The cooperative deceleration compensation amount is superimposed on the main drive control loop and combined with the target feedforward traction force command to generate the machining process control command with physical constraints.

[0014] Furthermore, the processing control commands output by the feedforward control module include annealing compensation commands, and the processing control commands with physical constraints are output to the actuator, specifically for: The number of delay cycles required for the wire to cross the physical distance between the forming zone and the annealing heating zone is calculated by combining the traction displacement; the annealing compensation instruction included in the processing control instruction is written into the first-in-first-out queue, and the read pointer offset of the bottom queue is continuously monitored; when the read pointer offset reaches the number of delay cycles, the corresponding annealing compensation instruction is removed from the first-in-first-out queue and sent to the actuator.

[0015] Furthermore, the parameter optimization module is used to perform abnormal sample rejection and preprocessing based on the end-product quality inspection data corresponding to the processing control instructions, specifically for: Abnormal sample removal and preprocessing are performed on the quality inspection data of the terminal products, and valid samples are retained. The feedforward control gain parameter update amount is calculated using the parameter incremental optimization formula based on the deviation between the actual size and the target size corresponding to the valid sample. The feedforward control gain parameter update amount is added to the gain coefficient of the previous production batch to obtain the preliminary updated gain parameters.

[0016] Furthermore, after obtaining the initially updated gain parameters, the parameter optimization module, after performing limit boundary checks and safety verification, outputs the updated global control gain parameters, specifically used for: Limit boundary checks are performed on the initially updated gain parameters. If the values ​​exceed the set boundaries, the initially updated gain parameters are forcibly clamped to the corresponding boundary values. Based on the boundary values ​​after the forced clamping or the initially updated gain parameters that do not exceed the limits, a safety review prompt is triggered on the human-machine interface. After the safety review prompt is inspected and approved by the process personnel, the data is packaged and the updated global control gain parameters are output.

[0017] This invention provides an intelligent control system for processing parameters in the copper busbar manufacturing process. It has the following beneficial effects: 1. This invention converts continuous time-dimensional thermal state data into discrete spatial-dimensional material slice data through a data mapping module, and constructs a material coordinate mapping table by combining reverse addressing logic. This technical feature solves the delay problem caused by physical displacement between processes in the production line, enabling the control system to track the specific cooling state of the material entering the forming zone, eliminating the impact of spatial state mismatch, and providing data support for position matching for applying traction force compensation.

[0018] 2. This invention utilizes a state decoupling module to perform decoupling calculations on the field electrical feedback parameters. It subtracts the no-load current component for overcoming mechanical friction and the inertial compensation current component during acceleration and deceleration from the total system current, extracting the equivalent forming load current. Simultaneously, an external interference shielding mechanism is established to prevent abnormal data generated during sudden equipment stops or jams from interfering with wear condition assessment. This feature effectively separates the inherent resistance of the mechanical system from the actual wear state of the forming die, eliminating parameter coupling deviations under multiple operating conditions and providing the control system with objective quantitative indicators of die wear.

[0019] 3. This invention constructs a linkage control mechanism based on feedforward control and parameter optimization. Based on the acquired cooling status and wear index, it outputs traction force commands with physical limitations in advance, intervening before sudden changes in resistance and compensating for the shortcomings of conventional delayed feedback. For overload conditions, the system sets a safe upper limit threshold for traction force and uses a combination of traction force clamping and coordinated deceleration to prevent system overload. The underlying layer uses a first-in-first-out queue to match the delay period of the mechanical span, ensuring that the annealing compensation command is accurately applied in space. Simultaneously, the parameter optimization module performs closed-loop iterative updates to the global control gain parameters based on the quality inspection data of the end product, enabling the control system to adapt to the long-term aging characteristics of the equipment, preventing traction force instability and material breakage during operation, and reducing the product dimensional deviation rate. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the architecture of the intelligent control system for processing parameters in the copper busbar manufacturing process of the present invention; Figure 2 This is a flowchart of the intelligent control method for processing parameters in the copper busbar manufacturing process of the present invention; Figure 3 This is a data flow architecture diagram for the thermal state archiving during the upward continuous casting rod production stage of the present invention; Figure 4 This is a diagram of the cross-process material space coordinate mapping architecture of the present invention; Figure 5 This is a diagram of the main drive load soft measurement and state identification control architecture of the present invention; Figure 6 This is a diagram of the boundary constraint-driven collaborative feedforward control architecture of the present invention; Figure 7 This is a diagram of the offline parameter iterative optimization architecture based on batch data of the present invention; Figure 8 This is a diagram showing the traction force fluctuation distribution of the conventional control scheme of the present invention; Figure 9 This is a diagram showing the traction force fluctuation distribution of the feedforward control scheme of the present invention. Detailed Implementation

[0021] 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.

[0022] See attached document Figure 1 This invention provides an intelligent control system for processing parameters in copper busbar manufacturing. The system comprises, in terms of hardware, data acquisition equipment, a programmable logic controller network, an industrial communication bus, actuators, and a host computer database server.

[0023] Among them, the data acquisition equipment is respectively deployed at the site of the continuous casting rod making process, the continuous extrusion or rolling process, the drawing or stretching shaping process, and the annealing process, to collect the state parameters in the production process and form on-site working condition data; Programmable logic controller (PLC) networks are used to receive field operating data and combine them with global control gain parameters to perform logic operations and timing control, generating process control instructions. Industrial communication buses are used to transmit process control commands and cross-process data to form digital communication links between equipment in each process. The actuator receives process control commands through a digital communication link, adjusts the on-site process status, and feeds back execution status parameters; The host computer database server is used to record execution status parameters and historical cross-process data, and output global control gain parameters for subsequent iteration updates.

[0024] In terms of functional division, the system includes a data mapping module, a state decoupling module, a feedforward control module, and a parameter optimization module. The data mapping module, state decoupling module, and feedforward control module are configured within the programmable logic controller network; the parameter optimization module is configured within the host computer database server.

[0025] The data mapping module is used to receive on-site operating condition data, convert continuously collected thermal state data into discrete material slice data, monitor the material transfer status, and output a comprehensive cooling intensity index and a material coordinate mapping table. The state decoupling module is used to receive on-site working condition data and process the electrical feedback parameters therein to extract the equivalent forming load current or equivalent forming load torque, calculate and output the equivalent wear correlation index of the forming die; The feedforward control module combines the comprehensive cooling intensity index, material coordinate mapping table, equivalent wear correlation index of forming die, and global control gain parameters to output machining process control commands with physical constraints to the actuator. These machining process control commands include at least one of traction force control commands, running speed control commands, and annealing heating compensation commands. The global control gain parameters include traction force compensation gain parameters, and when the system is configured in annealing compensation adaptive update mode, it also includes annealing compensation gain parameters. The parameter optimization module is used to perform a safety review based on the end-product quality inspection data corresponding to the processing control instructions after the batch production cycle is completed, and output the updated global control gain parameters.

[0026] See attached document Figure 2 The present invention also provides an intelligent control method for processing parameters in the copper busbar manufacturing process, the method comprising the following steps: S1, Thermal status filing of the upper continuous casting rod production stage: Obtain on-site working condition data, divide virtual material slices according to fixed length, and calculate the comprehensive cooling intensity index corresponding to each virtual material slice based on temperature and speed data; S2, cross-process material space coordinate mapping: after the coil or billet is transferred from the upper continuous casting rod making process to the continuous extrusion, rolling or drawing forming process, the material space coordinate is updated according to the physical inversion characteristics and length loss events in the material receiving and unloading process, and the material coordinate mapping table is output in combination with the comprehensive cooling intensity index. S3, Main drive load soft measurement and status identification: Acquire field working condition data, process the current data, speed data or torque estimation data fed back from the main drive of continuous extrusion, rolling or drawing and forming, extract the equivalent forming load current or equivalent forming load torque through multi-working condition decoupling calculation, and calculate and output the current forming die equivalent wear correlation index. S4, Boundary Constraint-Driven Collaborative Feedforward Control: The feedforward target is calculated based on the comprehensive cooling intensity index, material coordinate mapping table, forming die equivalent wear correlation index, and global control gain parameters; when the feedforward target reaches the set safety threshold, the traction force protection mechanism is triggered to limit the feedforward traction force command within the safe amplitude range, and collaborative deceleration control is executed to output the processing control command. S5, offline parameter iteration based on batch data: Based on the terminal product quality inspection data corresponding to the processing control instructions, after removing discrete abnormal samples, calculate the update amount of the feedforward control gain parameter, perform limit boundary verification on the update result, and output the iterated global control gain parameter.

[0027] The following section, in conjunction with each functional module, further explains the principles of physical mapping, state identification, and collaborative control of processing parameters involved in the system operation process in this embodiment.

[0028] See attached document Figure 3 In the upward continuous casting rod-making process, copper rods or copper billets are produced continuously. To facilitate subsequent tracking and retrieval of continuous materials, the system converts the continuously collected thermal data into virtual slice data divided by length. Step S1 above, namely the thermal state filing for the upward continuous casting rod-making stage, specifically includes the following steps: S101, obtain the real-time physical parameters of the upward continuous casting holding furnace and the crystallizer.

[0029] In this embodiment, the data acquisition device includes a temperature sensor installed inside the holding furnace, an encoder installed at the outlet end of the crystallizer, and a flow sensor and a temperature sensor installed on the cooling water pipe. As a specific implementation, the temperature sensor can be a PT100 resistance temperature detector (RTD), and the encoder can be an Omron E6B2 series rotary encoder.

[0030] The data acquisition device converts the acquired analog or pulse signals into digital data that can be processed by the controller, and outputs the digital data as field operating data to the data mapping module in the programmable logic controller network. Regarding the installation locations of the aforementioned sensors and encoders and the signal conversion process, those skilled in the art can perform conventional configurations according to the field environment. The hardware selection and signal conversion principles are well-known technologies in this field and will not be elaborated upon here.

[0031] S102, Construct a virtual material slice sequence.

[0032] To discretize the continuously produced copper rods or billets, the data mapping module sets a fixed length parameter, defined as the unit slice length. This unit slice length is input by the field process personnel through the system configuration interface, based on control precision and data storage capacity. As a preferred approach, the unit slice length ranges from 1m to 5m.

[0033] The system converts the cumulative pulse count fed back by the encoder into the actual traction length based on the circumference of the traction wheel of the meter counter and the number of pulses per revolution. Then, it divides the actual traction length by the unit slice length, thereby logically dividing the continuous copper rod or copper billet into virtual slices of equal length. After the division is completed, the data mapping module generates a set of virtual slices and assigns a unique virtual slice number to each virtual slice according to the production sequence.

[0034] S103, Construct a normalized model.

[0035] In this embodiment, the furnace temperature in the holding furnace is used to characterize the initial thermal state of the molten copper before it enters the crystallization stage; the relationship between the temperature difference between the inlet and outlet cooling water of the crystallizer and the traction speed is used to characterize the strength of the cooling effect on a unit length of copper rod or copper billet during the crystallization stage. Since temperature, temperature difference, and speed are process parameters with different dimensions, the data mapping module uses a normalized formula to calculate the comprehensive cooling intensity index of each virtual slice.

[0036] Specifically, for a single virtual slice, the system calculates the first... The normalized formula for the comprehensive cooling intensity index of each virtual slice is: ; In the formula, Indicates the first The overall cooling intensity index of a virtual slice; This represents the first preset weight allocation coefficient; Indicates the first The temperature difference between the inlet and outlet cooling water when a virtual slice passes through the crystallizer; This indicates the reference cooling water inlet and outlet temperature difference under standard process conditions; Indicates the first The traction speed of a virtual slice as it passes through the crystallizer; This indicates the reference traction speed under standard process conditions; This represents the second preset weight allocation coefficient; Indicates the first The furnace temperature when a virtual slice passes through the holding furnace; This indicates the reference furnace temperature under standard processes.

[0037] in, This represents the dimensionless deviation rate of the actual cooling water temperature difference from the reference cooling water temperature difference. This represents the traction speed ratio; dividing the two yields the relative change in cooling temperature difference compensated for by the traction speed. This indicates the deviation between the actual furnace temperature and the reference furnace temperature. This indicates the relative rate of change of furnace temperature.

[0038] Among the aforementioned benchmark parameters, the benchmark cooling water inlet and outlet temperature difference, benchmark traction speed, and benchmark furnace temperature under standard processes are all obtained by the data mapping module from the process instruction data dictionary of the current production batch, read from the host computer database server. The values ​​of the first preset weight allocation coefficient and the second preset weight allocation coefficient are both greater than 0 and less than 1, and their sum is 1. Their specific values ​​can be obtained by process engineers through fitting and calibration based on historical normal batch production data.

[0039] To avoid calculation anomalies caused by division by zero or invalid baseline parameters during normalization, the data mapping module performs a validity check before substituting the above normalization formula. If the first... The traction speed of a virtual slice as it passes through the crystallizer The value is 0, or the reference cooling water inlet and outlet temperature difference under standard process conditions is detected. Less than or equal to 0, or a reference traction speed under standard process conditions is detected. Less than or equal to 0, or the reference furnace temperature under standard process conditions is detected. If the value is less than or equal to 0, then the current data is determined not to meet the normalization calculation conditions.

[0040] In the above scenario, the data mapping module does not perform the normalization operation. Instead, it assigns a value of 0 to the comprehensive cooling intensity index of the corresponding virtual slice, or uses the comprehensive cooling intensity index from the previous stable sampling period. Through this processing method, the data mapping module converts the crystallization cooling state corresponding to the virtual slice and its deviation from the standard process state into a unified-dimensional digital feature for subsequent cross-process mapping and feedforward control.

[0041] S104, Cross-process historical data binding and storage archiving.

[0042] Once the current production batch of coils or blanks has been wound, transferred, or registered in the warehouse, the data mapping module acquires the physical identifier scanned and read by the industrial identification terminal. As a specific implementation, the industrial identification terminal can employ radio frequency identification (RFID) reading devices or optical identification devices. The tag binding and data reading methods are well-known technologies in the field and will not be elaborated upon here.

[0043] The data mapping module establishes a data matrix table corresponding to the current coil or billet based on the physical identifier. This data matrix table includes the physical identifier of the coil or billet, the set of virtual slices, and the comprehensive cooling intensity index corresponding to each virtual slice number.

[0044] To enable cross-process data flow, the data matrix table, serving as the foundation for the material coordinate mapping table, is transmitted to the host computer database server via an industrial communication bus. The host computer database server stores the received data matrix table in its internal memory, thereby completing the archiving of thermal state data for the upward continuous casting rod stage. This provides a data basis for recalling historical thermal states based on material location in subsequent continuous extrusion, rolling, drawing, stretching or shaping, and annealing processes.

[0045] See attached document Figure 4 After the coil or billet is transferred from the continuous casting rod-making process to the continuous extrusion, rolling, or drawing forming process, the material may exhibit a spatial inversion relationship of first-in, last-out during the coiling, unloading, or feeding processes. To ensure that the virtual slice data established in the continuous casting rod-making stage corresponds to the actual feeding position in subsequent processes, the system remaps the physical coordinates of the material. Step S2 above, namely cross-process material spatial coordinate mapping, specifically includes the following steps: S201, Analysis of Physical Inversion Characteristics of Material Feeding and Discharging and Initial Coordinate Mapping.

[0046] In this embodiment, the data acquisition device reads the material roll or billet identification data collected by the industrial identification terminal at the feeding end of the subsequent processing step to obtain the physical identifier of the currently online material. As a preferred implementation, the industrial identification terminal can be a radio frequency identification (RFID) reader or an optical reader.

[0047] After obtaining the physical identifier, the data mapping module in the programmable logic controller network sends a query request to the host computer database server via the industrial communication bus to retrieve the data matrix table established during the upper continuous casting rod production stage.

[0048] During the winding process of the continuously cast rod, the material produced first is located in the inner layer of the coil, and the material produced later is located in the outer layer. When the coil enters the subsequent processing steps and begins to be unloaded, the outer layer material enters the processing equipment first. Therefore, the data mapping module reverses the sequence of virtual slice numbers in the data matrix table to form an initial logical mapping sequence consistent with the actual unloading order. If the current material is transferred as a fixed-length billet and there is no reversed winding / unloading relationship, the data mapping module establishes an initial logical mapping sequence according to the actual feeding direction of the billet.

[0049] S202, Slice offset compensation for field length loss events.

[0050] In this embodiment, during the processes of coil transfer, end treatment, continuous extrusion feeding, rolling bite-in, and drawing or stretching shaping, there may be trimming operations involving oxidized sections, damaged sections, or trial run scrap sections. These trimming operations alter the length reference of the material entering the forming zone. If the original slice sequence is still used for addressing, the thermal state data recorded during the upward continuous casting rod stage will be inconsistent with the actual material position entering the forming zone.

[0051] To eliminate this positional deviation, the data mapping module combines the meter counter feedback value from the on-site working conditions with the shearing length entered by the process personnel through the human-machine interface to calculate the slice offset of the current material position. The data mapping module uses a slice index addressing formula to calculate the target virtual slice number entering the forming area at the current moment. The slice index addressing formula is: ; In the formula, This indicates the target virtual slice number that has entered the forming area at the current moment; Indicates the current sampling time; This indicates the total number of virtual slices contained in the current roll or billet; This indicates the cumulative feed length of the processing equipment at the current sampling time; This indicates the cumulative length loss of the current coil or billet; Indicates the unit slice length; This indicates the floor function.

[0052] in, This indicates the total length of material consumed at the current sampling time. This represents the theoretical total number of slices consumed at the current sampling time.

[0053] All length parameters mentioned above are in meters (m). This refers to the total number of virtual slices contained in the current roll or billet. Reading the dimension attributes of the data matrix table issued by the host computer database server; the cumulative feed length of the processing equipment at the current sampling time. The cumulative length loss of the current coil or billet is calculated by the data mapping module based on the pulse signals from the metering wheel at the feed end. It is obtained by summing up the on-site recorded values ​​such as end shear length and test waste length.

[0054] To prevent computational overflow caused by abnormal unit slice lengths, the system configures the unit slice length during the parameter configuration phase. Set a lower limit protection to ensure it is always greater than 0. When When the value is exactly an integer multiple of the unit slice length, the system determines the target slice number according to the leading edge position of the forming area in the current sampling period, and ensures the uniqueness of the addressing result through subsequent boundary constraint logic.

[0055] Therefore, the data mapping module converts the loss offset on the continuous length coordinate into the index offset on the discrete slice number, realizing coordinate compensation for length loss events such as trimming and trial waste.

[0056] S203, Target slice boundary addressing and mapping table update based on boundary constraints.

[0057] After calculating the target virtual slice number that enters the forming area at the current moment, the data mapping module performs boundary constraint verification on the target virtual slice number to avoid array out-of-bounds errors caused by sensor errors, metering anomalies, or manual input errors.

[0058] If the target virtual slice number entering the forming area at the current moment is less than 1, the data mapping module clamps it to 1; if the target virtual slice number is greater than the total number of virtual slices contained in the current roll or billet, the data mapping module triggers a coordinate anomaly alarm and freezes the current addressing coordinates.

[0059] After boundary verification, the data mapping module extracts the corresponding comprehensive cooling intensity index from the data matrix table based on the target virtual slice number. Subsequently, the data mapping module binds the current physical location coordinates of the processing equipment, the target virtual slice number, and the corresponding comprehensive cooling intensity index, updating and outputting the material coordinate mapping table. This material coordinate mapping table serves as the spatial reference for subsequent state identification and feedforward control, and is used by the state decoupling module and the feedforward control module.

[0060] See attached document Figure 5 When continuous extrusion, rolling, or drawing forming equipment is running, the current or torque feedback of the main drive motor reflects not only the forming resistance of the copper busbar billet, but also the load component introduced by mechanical friction and acceleration / deceleration inertia. To reduce the influence of the above non-forming factors on the determination of the die or roll state, the system extracts the equivalent load information related to the forming resistance through a state decoupling module. Step S3 above, namely the main drive load soft measurement and state identification, specifically includes the following steps: S301, Decoupling and extraction of equivalent shaped load current under multi-condition coupling.

[0061] In this embodiment, the data acquisition device obtains the operating parameters of the main drive motor for continuous extrusion, rolling, or drawing forming, and transmits the field operating condition data containing the above parameters to the state decoupling module in the programmable logic controller network. As a preferred embodiment, the main drive system is driven by a frequency converter or servo driver, and the state decoupling module processes the current data, speed data, or torque estimation data fed back from the main drive.

[0062] The current signal fed back from the main drive typically includes a load current component to overcome the forming resistance of the copper busbar blank, an unloaded current component to overcome the frictional resistance of the mechanical system, and an inertia compensation current component during acceleration and deceleration. Therefore, the state decoupling module performs multi-condition decoupling calculations to extract the equivalent forming load current.

[0063] To avoid mixing the total current feedback from the driver with the torque current component, the state decoupling module defines the equivalent drive current. This serves as the input to the current decoupling model. The equivalent drive current can be obtained by on-site calibration and conversion from the total current fed back by the main drive inverter, or it can be directly obtained from the torque current component output by the drive.

[0064] To achieve the above decoupling process, the state decoupling module uses the current decoupling calculation formula to calculate and extract the equivalent shaped load current. The current decoupling calculation formula is as follows: ; In the formula, This represents the equivalent shaped load current at the current sampling moment; Indicates the current sampling time; This represents the equivalent driving current at the current sampling moment; This represents the reference value of the no-load current at the current speed; This represents the motor angular velocity at the current sampling moment; This represents the equivalent moment of inertia of the main drive system referred to the motor shaft side; This represents the motor torque constant; This represents the angular acceleration at the current sampling moment.

[0065] As a preferred method, angular acceleration is obtained by the difference in motor angular velocity between adjacent sampling periods, and the calculation relationship is as follows: ; in, This indicates the sampling period of the state decoupling module.

[0066] In the above current decoupling calculation formula, This represents the inertial compensation torque required for dynamic acceleration. This represents the inertial compensation current derived from the inertial compensation torque. This is achieved by introducing the motor torque constant. The inertial compensation term and the current term maintain the same dimensions.

[0067] Regarding the acquisition method of the above parameters, the equivalent driving current at the current sampling moment... The motor angular velocity at the current sampling moment is obtained from the communication messages of the frequency converter or servo driver, torque current components, or total current calibration conversion results; the no-load current reference value at the current speed is obtained from the speed-current lookup table mapping relationship established during the no-load test run of the equipment; the equivalent moment of inertia converted to the motor shaft side is obtained from the motor self-learning identification program built into the driver, mechanical transmission parameter conversion, or on-site debugging calibration; the motor torque constant is obtained from the motor nameplate parameters, driver parameters, or on-site identification results.

[0068] Therefore, the state decoupling module can obtain the equivalent forming load current, which characterizes the forming resistance of the copper busbar blank, after deducting the no-load friction component and the inertial compensation component.

[0069] As an alternative implementation, the state decoupling module can also directly read the estimated torque output by the driver and decouple it in torque form: ; in, This represents the equivalent forming load torque at the current sampling moment; This indicates the total output torque, either from driver feedback or estimation. This represents the reference value of the no-load torque at the current speed.

[0070] When calculating in torque form, the state decoupling module uses the following relationship to calculate the equivalent wear correlation index of the forming die: ; In the formula, This represents the baseline equivalent forming load torque under new mold, roll, or forming fixture conditions. If detected... If the value is less than or equal to 0, the state decoupling module will stop performing the exponential calculation of the torque form and issue a prompt to initialize the reference parameters.

[0071] S302, Calculation of Equivalent Wear Correlation Index and Setting of Safety Boundaries for Forming Dies.

[0072] As the continuous extrusion, rolling, or drawing process continues, the forming die, rolls, or drawing fixtures may wear, causing changes in contact friction and forming zone resistance. This change is typically reflected in the increasing trend of the equivalent forming load current or equivalent forming load torque. Based on this physical correspondence, the state decoupling module, after extracting the equivalent forming load current or equivalent forming load torque, calculates and outputs the equivalent wear correlation index of the forming die.

[0073] When using the equivalent forming load current for calculation, the wear condition assessment formula is: ; In the formula, This represents the equivalent wear correlation index of the forming die; This represents the equivalent shaped load current at the current sampling moment; This represents the baseline equivalent forming load current under the condition of a brand-new mold, roll, or forming fixture. This formula is applicable when the state decoupling module uses the equivalent drive current for state identification. When the state decoupling module uses the equivalent forming load torque for state identification, the equivalent wear correlation index of the forming mold is obtained according to the calculation relationship of the aforementioned torque form.

[0074] in, This represents the equivalent forming load current increment. The forming die equivalent wear correlation index is used to characterize the load increase trend caused by the combined effects of die wear, roll wear, changes in friction state, and changes in material forming resistance. It is not limited to the direct measurement results of the geometric wear of the die or roll.

[0075] In this embodiment, the reference equivalent forming load current under the condition of a brand-new mold, roll, or forming fixture is automatically sampled and averaged by the state decoupling module during steady-state operation after each change of fixture, taking into account the current copper busbar specifications, current processing pass, and rated stable operating speed. The average value is then stored in an internal register as a reference value. If there is a difference between the actual operating speed and the reference sampling speed, the state decoupling module can correct the reference equivalent forming load current according to a pre-established speed compensation table. The method of establishing the speed compensation table is a conventional calibration method in the art and will not be described in detail here.

[0076] To prevent division errors caused by abnormal reference parameters, the state decoupling module performs a numerical validity check before substituting the values ​​into the wear condition assessment formula. If the reference equivalent forming load current is less than or equal to 0 in the condition of a brand-new mold, roll, or forming fixture, it is determined that the reference parameters have not been effectively initialized. The state decoupling module then stops executing the current exponent calculation and issues a parameter initialization prompt.

[0077] In addition, to ensure production safety, process engineers set a safety boundary threshold for the equivalent wear correlation index in the host computer database server. As a standard safety setting, this safety boundary threshold is typically set between 0.15 and 0.20. The state decoupling module cyclically compares the calculated equivalent wear correlation index of the forming die with this threshold; if the detected value exceeds the set safety boundary, a tooling replacement or maintenance warning signal is triggered through the industrial control network.

[0078] S303, Wear state freezing strategy under abnormal external interference.

[0079] In copper busbar production, unsteady conditions such as sudden equipment stoppage, billet jamming, abnormal lubrication, rolling slippage, abnormal drawing, or external mechanical impact may cause sudden changes in the main drive current signal or torque estimation signal. If state decoupling and index evaluation are continued under such conditions, the equivalent wear correlation index of the forming die is likely to deviate from the true stable operating state, thereby affecting the generation of subsequent feedforward control commands.

[0080] To reduce the impact of the above-mentioned abnormal operating conditions on the state identification results, the state decoupling module is equipped with a state freezing mechanism.

[0081] To implement this state freeze mechanism, the state decoupling module monitors the angular acceleration, stuck discrete signal, lubrication pressure, lubrication flow rate, and emergency stop status signal at the current sampling time in real time. When the absolute value of the angular acceleration at the current sampling time exceeds the preset allowable acceleration limit, or when the field operating data contains abnormal discrete signals fed back by the stuck sensor, or when external abnormal signals such as lubrication pressure, lubrication flow rate, or emergency stop status are detected, the state decoupling module determines that the current system is in an abnormal state of external interference.

[0082] Those skilled in the art will understand that the aforementioned preset upper limit of acceleration is calculated from the rated output torque and load inertia of the drive system, and is usually set to 1.2 to 1.5 times the normal production rated angular acceleration.

[0083] Under abnormal external interference conditions, the state decoupling module stops executing the aforementioned current decoupling calculation formula, torque decoupling calculation formula, and wear state assessment formula, and maintains the equivalent wear correlation index of the forming die at the historical calculation value of the previous stable operating cycle. After the external abnormal signal is eliminated and the drive system returns to steady-state operation, the state decoupling module unfreezes the state, restores real-time state identification, and outputs the equivalent wear correlation index of the forming die for the feedforward control module to perform traction force or forming load compensation control.

[0084] See attached document Figure 6 In the copper busbar production process, the thermal state differences formed during the upward continuous casting stage can cause changes in the local forming resistance of the material. Wear of the forming die, rolls, or drawing tooling can further increase the load in the forming zone. The combined effect of these factors can easily lead to fluctuations in traction force or forming load. To compensate for such disturbances within the safety boundaries of equipment and materials, the system performs coordinated control of traction force, speed, and annealing compensation through a feedforward control module. Step S4 above, namely the boundary constraint-driven coordinated feedforward control, specifically includes the following steps: S401, Feedforward target measurement based on cross-process feature mapping.

[0085] In this embodiment, the feedforward control module locates the physical slice currently entering the forming zone in the material coordinate mapping table based on the meter counter feedback value in the field working condition data, and extracts the comprehensive cooling intensity index corresponding to the physical slice.

[0086] From a technological perspective, slices with a higher comprehensive cooling intensity index typically exhibit higher forming resistance. When the equivalent wear correlation index of the forming die is high, the impact of the friction state of the die, rolls, or tooling on the traction load also increases accordingly. Therefore, the feedforward control module calculates the feedforward target by combining the comprehensive cooling intensity index, the material coordinate mapping table, the equivalent wear correlation index of the forming die, and the global control gain parameters.

[0087] The feedforward control module uses the feedforward traction force calculation formula to calculate the initial feedforward traction force command at the current sampling time. The feedforward traction force calculation formula is as follows: ; In the formula, This indicates the initial feedforward traction command at the current sampling moment; Indicates the current sampling time; Indicates the reference traction force setting value; This represents the first traction compensation gain coefficient in the global control gain parameters; This represents the comprehensive cooling intensity index corresponding to the target virtual slice that enters the forming zone at the current sampling moment; This represents the second traction force compensation gain coefficient in the global control gain parameters; This represents the equivalent wear correlation index of the forming mold at the current sampling time.

[0088] in, Indicates the amount of traction compensation under thermal conditions; It indicates the amount of traction force compensation caused by wear and friction changes in the mold, rolls, or tooling.

[0089] Regarding the acquisition methods of the above parameters, the baseline traction force setting value is directly set by process engineers based on the current copper busbar material, cross-sectional dimensions, and processing passes, referring to the process standard manual. The first traction force compensation gain coefficient and the second traction force compensation gain coefficient are the core components of the global control gain parameters. Their initial values ​​are set by engineering commissioning personnel based on on-site trial experience, and subsequently iteratively updated by the upper computer database server through the parameter optimization module after the batch production cycle ends. For conventional copper busbar drawing, stretching and shaping, or continuous forming processes, the first traction force compensation gain coefficient and the second traction force compensation gain coefficient can be determined according to the traction force unit, copper busbar specifications, processing passes, equipment transmission ratio, and on-site calibration results; as an example, they can be selected within the range of 50N to 300N.

[0090] By calling the aforementioned feedforward traction force calculation formula, the feedforward control module obtains a comprehensive traction force command target that simultaneously reflects the thermal state of the preceding stage and the current load state of the forming tooling.

[0091] S402, Traction clamping and deceleration coordination mechanism under severe tooling wear.

[0092] During continuous production, when the wear of molds, rolls, or forming fixtures intensifies or the friction condition significantly deteriorates, simply increasing the traction force compensation may cause the traction load on the copper busbar billet to approach or exceed the safe range, resulting in surface defects, dimensional instability, or billet jamming. Therefore, as a preferred approach, a safe upper limit threshold for traction force is set within the feedforward control module. This safe upper limit threshold is set by process engineers based on the copper busbar material, cross-sectional dimensions, allowable deformation load, and equipment safety margin, and is pre-stored in the host computer database server.

[0093] The feedforward control module compares the initial feedforward traction force command at the current sampling moment with the traction force safety upper limit threshold. When the initial feedforward traction force command at the current sampling moment is greater than or equal to the traction force safety upper limit threshold, the feedforward control module triggers the traction force protection mechanism, limiting the actual output traction force command to the traction force safety upper limit threshold, and generating the target feedforward traction force command.

[0094] After the traction force is clamped, the system no longer increases the traction force output, but instead reduces the forming load by decreasing the material running speed. The feedforward control module uses the cooperative deceleration calculation formula to calculate the cooperative deceleration compensation amount at the current sampling moment. The cooperative deceleration calculation formula is as follows: ; In the formula, This represents the amount of coordinated deceleration compensation at the current sampling moment; Indicates the speed compensation coefficient; This indicates the initial feedforward traction command at the current sampling moment; This indicates the upper limit threshold for traction safety.

[0095] in, This indicates the difference in traction force exceeding the limit.

[0096] Furthermore, the feedforward control module calculates the actual target speed command based on the cooperative deceleration compensation amount. The actual target speed command satisfies the following constraints: ; In the formula, This indicates the actual target velocity command at the current sampling moment; Indicates the minimum operating speed threshold; This indicates the base operating speed at the current sampling moment; This indicates the amount of coordinated deceleration compensation at the current sampling moment.

[0097] Those skilled in the art will understand that the speed compensation coefficient It can be calculated or calibrated from the transmission ratio of the drive system, the parameters of the reducer, the number of processing passes and the results of on-site debugging, and is used to represent the speed correction range corresponding to the unit traction force exceeding the limit.

[0098] To prevent the deceleration command from exceeding the allowable operating range of the drive system, the feedforward control module performs a minimum speed limit check before outputting the coordinated deceleration compensation. When the actual target speed after compensation is lower than the set minimum operating speed threshold, the system maintains the target speed at the minimum operating speed threshold and triggers a process degradation alarm. If the traction force exceeding the limit is not resolved within a preset duration, or if safety inputs such as stuck sensors or emergency stop buttons are triggered, the system further executes shutdown protection.

[0099] Through the above processing, the feedforward control module superimposes the coordinated deceleration compensation amount into the main drive control loop and issues a deceleration command to the actuator, realizing dynamic coordination between traction force and speed. Under normal operating conditions where no limit is exceeded, the feedforward control module assigns the target feedforward traction force command to the initial feedforward traction force command at the current sampling time, and assigns the coordinated deceleration compensation amount at the current sampling time to 0.

[0100] S403, traction force anti-sudden change output based on ramp function limitation.

[0101] After determining the target feedforward traction force command, the feedforward control module applies a ramp limit to the command to prevent abrupt changes in the traction force command from causing drive overload or mechanical shock. This ramp function generator limits the maximum rate of change of the command per unit time, thereby generating smoothly transitioning machining process control commands.

[0102] The internal algorithm implementation and parameter configuration of the ramp function generator, such as rise time and fall time, can be conventionally set by those skilled in the art based on the load response characteristics of the actuator. Such configuration methods are well-known in the field and will not be elaborated upon here. After processing, the feedforward control module issues machining process control commands with physical constraints to the actuator.

[0103] S404, Annealing instruction delay queue issued based on material coordinate tracking.

[0104] Because of the physical distance between the forming zone and the annealing heating zone, the material needs a certain amount of time to travel from the forming equipment outlet to the annealing heating zone. To ensure that the annealing compensation command applies to the same material slice corresponding to the traction force compensation, the feedforward control module, after calculating the annealing compensation command for the current slice, does not immediately issue it, but instead writes the annealing compensation command into a memory-based first-in-first-out queue.

[0105] In this embodiment, the annealing compensation command includes at least one of the following: output current correction, output voltage correction, or output power correction of the annealing heating equipment. As a preferred embodiment, the feedforward control module uses the comprehensive cooling intensity index corresponding to the target virtual slice entering the forming zone at the current sampling time. The equivalent wear correlation index of the forming die at the current sampling time. and basic annealing power Calculate the annealing power compensation target: ; In the formula, This indicates the annealing power compensation target at the current sampling moment; Indicates the base annealing power; This represents the gain coefficient for thermal annealing compensation. This represents the annealing compensation gain coefficient, indicating the equivalent wear-related state of the forming die.

[0106] in, and It can be used as part of the global control gain parameters in batch offline iterations, or it can be stored in the host computer database server as a pre-calibrated fixed annealing compensation parameter. As another implementation, the annealing power compensation target can also be obtained through a pre-calibrated process lookup table relationship.

[0107] The feedforward control module calculates the number of delay cycles required for the material to cross the physical distance between the forming zone and the annealing heating zone based on the traction displacement fed back by the meter counter in the field operating data. As a preferred method, the number of delay cycles satisfies: ; In the formula, This indicates the number of delay periods corresponding to the current sampling time; Indicates the length of the material path between the forming zone and the annealing heating zone; This indicates the material movement speed at the current sampling moment; This indicates the control cycle of the feedforward control module; The symbol indicates rounding up.

[0108] To prevent division by zero or abnormally increased delay periods during equipment shutdown or extremely low-speed operation, the feedforward control module performs a lower limit check on the material running speed before calculating the delay period. If the material running speed at the current sampling time is detected... If the speed is less than or equal to the minimum effective speed threshold allowed by the equipment, the feedforward control module will suspend the execution of the above delay period calculation formula and freeze the read pointer offset of the first-in-first-out queue; the delay period calculation and queue reading will resume after the material running speed recovers to the effective range.

[0109] Under normal operating conditions, when the read pointer offset of the underlying queue reaches the number of delay cycles, the feedforward control module removes the corresponding annealing compensation instruction from the first-in-first-out queue and sends it as part of the processing control instructions to the annealing heating equipment in the actuator via the industrial communication bus. Thus, the annealing compensation instruction can maintain consistency with the corresponding material slice in both time series and material space coordinates.

[0110] See attached document Figure 7 To adapt to the drift in operating conditions caused by long-term equipment operation and to maintain the dimensional stability of the finished copper busbars, the system performs offline parameter updates after each batch of production is completed. Step S5 above, namely offline parameter iteration based on batch data, specifically includes the following steps: S501, Abnormal sample removal and preprocessing of terminal quality data.

[0111] After a single batch production cycle is completed, the parameter optimization module running on the host computer database server receives end-product quality inspection data fed back by the quality inspection equipment. This end-product quality inspection data includes the actual thickness, width, or straightness measurements of the finished copper busbars during the batch production process.

[0112] In industrial production sites, signal pulse interference from measuring equipment, or non-steady-state production sections such as material insertion, jointing, jamming, and restarting, may cause some dimensional measurement data to deviate from the stable production state. To avoid such abnormal data affecting subsequent parameter optimization, the parameter optimization module performs abnormal sample rejection and preprocessing on the quality inspection data of the end product.

[0113] Specifically, the parameter optimization module calculates the mean and standard deviation of all target size measurements in the current batch, and uses the Laida rule, i.e. the 3σ criterion, to remove discrete outlier samples that exceed the range of plus or minus three times the mean standard deviation, retaining valid samples for subsequent model calculations.

[0114] If the number of dimensional measurement samples in the current batch is lower than the preset minimum number of samples, or if the calculated standard deviation is 0 and the sample values ​​lack effective changes, the parameter optimization module determines that the quality data of this batch does not meet the parameter iteration conditions, stops the execution of this abnormal sample removal and subsequent parameter iteration, and maintains the global control gain parameter of the previous production batch unchanged.

[0115] For the outlier removal process based on Laida's rule, those skilled in the art can perform conventional configurations according to mathematical statistics theory. The algorithm implementation is a well-known technology in this field and will not be described in detail here.

[0116] S502, incremental offline optimization of control gain parameter iterative update.

[0117] After sample preprocessing, the parameter optimization module calculates the update amount of the feedforward control gain parameter based on the deviation between the actual size and the target size corresponding to the valid samples. This calculation is used to determine whether the existing traction compensation gain needs to be adjusted under different thermal conditions and mold load conditions.

[0118] During the forming process, if a material slice under conditions of high overall cooling intensity or high mold wear continuously deviates from the target tolerance zone, it indicates that the traction compensation gain currently allocated to the corresponding state feature needs to be corrected. The parameter optimization module calculates the gain update amount along the error convergence direction based on the correspondence between dimensional deviations and state features.

[0119] Based on the above error feedback relationship, the parameter optimization module uses the parameter incremental optimization formula to calculate the increment. The parameter incremental optimization formula is as follows: ; ; In the formula, This indicates the update amount of the first traction force compensation gain coefficient; Indicates the first update direction coefficient; This represents the first learning rate parameter; This indicates the total number of valid samples in the current batch; Indicates the target set size; Indicates the first The actual measured dimensions of each valid sample; Indicates the first Normalized size deviation of each valid sample; Indicates the first The comprehensive cooling intensity index corresponding to each valid sample; This indicates the update amount of the second traction force compensation gain coefficient; Indicates the second update direction coefficient; This represents the second learning rate parameter; Indicates the first The equivalent wear correlation index of forming molds corresponding to each valid sample; Indicates all Summation is performed on each valid sample.

[0120] First update direction coefficient Second update direction coefficient Based on the on-site calibration results, its value is set to +1 or -1, which is used to determine whether the traction compensation gain increases or decreases along the error reduction direction, so as to adapt to the different responses of different equipment structures, processing steps and controlled objects to changes in traction force.

[0121] Before substituting the parameter incremental optimization formula, the parameter optimization module performs a valid check on the number of effective samples and the target size. If the total number of effective samples in the current batch... Equal to 0, or target set size If the value is less than or equal to 0, the parameter optimization module will stop executing the current iteration and assign the corresponding update value to 0 to avoid calculation errors caused by a denominator of 0.

[0122] After calculating the update amounts of the first traction compensation gain coefficient and the second traction compensation gain coefficient, the parameter optimization module adds them to the first traction compensation gain coefficient and the second traction compensation gain coefficient of the previous production batch to obtain the preliminary updated gain parameters.

[0123] In this embodiment, the parameter optimization module compensates for the gain coefficient with the first traction force. Second traction force compensation gain coefficient The primary iteration target; when the system is configured to synchronously optimize annealing compensation parameters, the parameter optimization module can also optimize based on post-annealing thickness fluctuations, width fluctuations, annealing temperature deviations, or resistivity deviations. and Perform offline updates under the same boundary constraints.

[0124] Regarding the first learning rate parameter Second learning rate parameter This is used to limit the parameter update step size. To avoid control fluctuations caused by excessive parameter changes between batches, the first learning rate parameter... Second learning rate parameter The value can be calculated based on 1% to 5% of the initial value of the corresponding traction force compensation gain coefficient. As for the specific calibration method of the learning rate parameter, those skilled in the art can perform conventional settings based on historical batch data and the convergence trend of dimensional errors, which will not be elaborated here.

[0125] S503, Limit Boundary Verification and Human-Computer Interaction Security Review Mechanism.

[0126] After obtaining the initially updated gain parameters, the parameter optimization module does not directly send them to the underlying controller, but first performs limit boundary checks. The host computer database server internally stores the parameter allowable range pre-defined by process engineers. To prevent the optimization results from deviating from the safe control range, the upper and lower limits of this parameter allowable range are usually set to ±20% to 30% of the initial set gain coefficient.

[0127] The parameter optimization module compares the initially updated first and second traction compensation gain coefficients with their corresponding upper and lower limits. If the values ​​exceed the set boundaries, they are forcibly clamped to the corresponding boundary values ​​to limit excessive correction of control parameters caused by abnormal data or occasional operating conditions.

[0128] After numerical verification is completed, the parameter optimization module triggers a security review prompt on the human-machine interface of the host computer database server, and displays a comparison trend curve before and after the parameter modification. Process personnel confirm or reject the operation through terminal inspection.

[0129] If the process engineer approves the parameter update, the parameter optimization module will package the confirmed parameters and output the updated global control gain parameters. If the process engineer refuses to modify the parameter update or fails to provide a confirmation response within the preset timeout period, the parameter optimization module will cancel the parameter update and maintain the parameter settings of the previous production batch.

[0130] Through the aforementioned boundary verification and manual review mechanisms, the system can correct the control gain using batch quality data while preventing unconfirmed parameter changes from affecting subsequent production safety. After successfully passing the safety review, the updated global control gain parameters are sent to the feedforward control module in the programmable logic controller network via the industrial communication bus before the start of the next production batch, for a new round of collaborative feedforward control.

[0131] Specific application examples: The following describes the operation process and comparative verification results of this embodiment, taking into account the production conditions of the T2 purple copper rectangular copper busbar production line of a copper processing plant.

[0132] In this embodiment, the system performs processing control on a batch of rectangular copper busbars with a nominal cross-sectional specification of 3.0 mm thickness. In the upper continuous casting rod process, the unit slice length is set to 2 m.

[0133] For virtual material slice #500, the data acquisition device obtains the furnace temperature at the time the slice is produced. The reference furnace temperature is 1125℃. The temperature is 1150℃; the temperature difference between the inlet and outlet of the cooling water. The reference cooling water inlet and outlet temperature difference is 8.2℃. Temperature: 8.0℃; Traction speed The reference traction speed is 2.1 m / min. The speed is 2.0 m / min. Set the weighting coefficient. , The data mapping module uses a normalized formula to calculate the overall cooling intensity index of the slice: ; The calculations above show that the virtual material slice #500 exhibits only a slight cooling deviation compared to the standard process conditions. This result can be used as the thermal state characteristic input for subsequent continuous extrusion, rolling, drawing, or stretching shaping stages. For production sections with larger deviations in cooling water temperature difference, traction speed, or furnace temperature, the comprehensive cooling intensity index increases accordingly. Figure 8 and Figure 9 The range of values ​​is used to cover different batches and different combinations of operating conditions under different disturbance levels.

[0134] After the material roll is transferred to the subsequent forming process, the meter counter provides feedback on the current cumulative feed length. The length is 995m. The length loss of this coil due to end oxidation during the initial trimming process is known. The value is 3.5m. The data mapping module uses the slice index addressing formula to calculate the target virtual slice number entering the forming area at the current moment: ; Assuming the total number of slices in this roll If the value is 1000, the addressing result points to the virtual material slice numbered #501, and the data mapping module retrieves the historical cooling characteristics corresponding to this physical location.

[0135] During the operation of the forming equipment, the state decoupling module reads the equivalent drive current from the data fed back by the actuator. The current is 85A. The reference value for no-load current at the current speed is 12A, the angular acceleration is 0, and the reference equivalent forming load current is... The current current is 60A. Since the current angular acceleration is 0, the inertial compensation current term is 0. The state decoupling module uses the current decoupling calculation formula to calculate the equivalent forming load current as 73A. Subsequently, the state decoupling module uses the wear state assessment formula to calculate the equivalent wear correlation index of the forming die: ; Since the value exceeds the preset safety threshold of 0.20, the system generates control commands and outputs a tooling replacement or maintenance warning signal to the human-machine interface.

[0136] The feedforward control module synthesizes commands based on the above parameters. It then sets the baseline traction force setpoint. The first traction force compensation gain coefficient is 1200N. The second traction force compensation gain coefficient is 180N. It is 250N. Because... and All are dimensionless state indices. and These represent the traction compensation amount corresponding to the unit thermal state index and the unit equivalent wear correlation index, respectively.

[0137] The feedforward control module uses the feedforward traction force calculation formula to calculate the initial feedforward traction force command: ; At this point, if the traction force safety upper limit threshold If the initial feedforward traction force command is set to 1250N, then the feedforward control module triggers the traction force protection mechanism, clamping the actual output traction force amplitude to 1250N, and calculates the deceleration compensation amount according to the cooperative deceleration calculation formula. Let the speed compensation coefficient be... ,but: ; If the current base operating speed is The minimum operating speed threshold is The system then further calculates the actual target speed command: ; Through the aforementioned traction clamping and coordinated speed reduction, the system limits the traction load overload of the copper busbar billet while restricting the operating speed to compensate for the operating fluctuations caused by the increase in forming resistance.

[0138] To evaluate the effect of this embodiment on suppressing process disturbances, a traditional PID control scheme was selected as a control group and compared with the cooperative feedforward control scheme of this embodiment. The verification data includes the distribution of traction force deviation under different combinations of comprehensive cooling intensity index and equivalent wear correlation index of the forming die.

[0139] The surface data used for illustration is obtained by fitting on-site calibration parameters and sampled data; in the case of incomplete on-site sampled data, simulation data based on on-site calibration parameters can also be used to generate comparative surfaces to characterize the changing trend of traction force deviation under different control strategies.

[0140] Figure 8 The results show that without cross-process feedforward compensation, the traction force deviation increases nonlinearly with the increase of the comprehensive cooling intensity index and the equivalent wear correlation index of the forming die, and the surface fluctuation is more obvious, indicating that the traditional control scheme is not stable enough in traction force under compound disturbance conditions.

[0141] Figure 9 The paper shows that after introducing the collaborative feedforward mechanism, the traction force deviation surface under the same working conditions is smoother than that of the traditional control scheme, and the deviation amplitude is reduced. This indicates that the system can provide feedforward compensation for cross-process thermal state disturbances and changes in the load state of forming tooling.

[0142] according to Figure 8 and Figure 9 The comparative data shows that the traditional control mode exhibits response lag when dealing with complex operating conditions. Under the indicated calibration parameters, when the comprehensive cooling intensity index exceeds 0.2 and the forming tooling enters the middle to late stage of wear (i.e., the equivalent wear correlation index of the forming die is greater than 0.15), the average traction force deviation of the traditional scheme exceeds 80N, accompanied by local pulse-like fluctuations. This phenomenon can easily lead to a decrease in the surface quality of the copper busbar and excessive deviations in cross-sectional dimensional consistency in actual production.

[0143] Comparative verification results based on on-site calibration parameters show that, under the same high cooling intensity and high wear correlation index conditions, the collaborative feedforward control scheme of this embodiment can control the traction force deviation within 15N. Combined with the real-time identification of the forming tooling load state by the state decoupling module, the feedforward control module can adjust the traction force and speed commands according to the cross-process state characteristics, thereby reducing the dimensional fluctuation of the finished copper busbar. As a verification result, in this verification batch, the standard deviation of the final finished copper busbar thickness can be reduced from 0.012mm under the traditional scheme to 0.003mm.

[0144] The aforementioned control effect stems from the correlation modeling between material coordinates, thermal characteristics, and forming load state. After the material coordinates and thermal characteristics are dynamically bound, forming control no longer adjusts solely based on single-point field feedback, but rather coordinates with the upstream thermal state and the current forming fixture load state for a collaborative response. Simultaneously, by introducing a collaborative deceleration calculation formula and a traction clamping mechanism, the system restricts and adjusts the operating speed during stages of severe wear on the mold, rolls, or forming fixtures, as well as deterioration of friction conditions, to reduce the risk of overload on the copper busbar billet, thereby reducing surface defects, dimensional instability, or jamming failures. Comprehensive comparative results show that the present invention improves the consistency of copper busbar dimensional accuracy while enhancing the operational stability of the production line.

[0145] 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. An intelligent control system for processing parameters in a copper busbar manufacturing process, characterized in that, include: The data mapping module is used to receive on-site operating condition data, convert continuously collected thermal state data into discrete material slice data, monitor the material transfer status, and output a comprehensive cooling intensity index and a material coordinate mapping table. The state decoupling module is used to receive the field working condition data, process the electrical feedback parameters in the field working condition data, extract the equivalent forming load current or equivalent forming load torque, and calculate and output the equivalent wear correlation index of the forming mold. The feedforward control module is used to combine the comprehensive cooling intensity index, the material coordinate mapping table, the forming die equivalent wear correlation index, and the global control gain parameter to output processing control commands with physical constraints to the actuator. The parameter optimization module is used to remove and preprocess abnormal samples based on the terminal product quality inspection data corresponding to the processing control instructions after the batch production cycle is completed. After performing limit boundary verification and safety review, it outputs the updated global control gain parameters.

2. The intelligent control system for processing parameters in a copper busbar manufacturing process according to claim 1, characterized in that, The data mapping module is used to receive the on-site operating condition data, convert the continuously collected thermal state data into discrete material slice data, and output the comprehensive cooling intensity index, specifically for: Based on the on-site working condition data, continuous copper rods or copper billets are divided into fixed lengths and a virtual slice set is generated. Using a normalized formula, the comprehensive cooling intensity index corresponding to each virtual slice is calculated based on the temperature and velocity data in the virtual slice set and the field operating condition data.

3. The intelligent control system for processing parameters in a copper busbar manufacturing process according to claim 1, characterized in that, The data mapping module is used to receive the on-site operating data, monitor the material transfer status, and, in conjunction with the output comprehensive cooling intensity index, output the material coordinate mapping table, specifically for: Based on the physical inversion characteristics during the take-up and release process, the virtual slice number sequence is reversed to form the initial logical mapping sequence; Based on the initial logical mapping sequence, and combined with the meter counter feedback value in the field working condition data and the recorded shearing length, the target virtual slice number entering the forming area at the current moment is calculated using the slice index addressing formula. Based on the target virtual slice number, the comprehensive cooling intensity index of the corresponding slice is extracted and bound, and the material coordinate mapping table is updated and output.

4. The intelligent control system for processing parameters in a copper busbar manufacturing process according to claim 1, characterized in that, The state decoupling module is used to receive the field operating condition data, process the electrical feedback parameters in the field operating condition data, extract the equivalent forming load current, calculate and output the equivalent wear correlation index of the forming mold, specifically for: The on-site working condition data is calculated using a current decoupling formula. The no-load current component that overcomes the frictional resistance of the mechanical system and the inertial compensation current component during acceleration and deceleration are deducted, and the equivalent forming load current is extracted. The wear condition assessment formula is used to calculate the current equivalent wear correlation index of the forming die based on the equivalent forming load current.

5. The intelligent control system for processing parameters in a copper busbar manufacturing process according to claim 4, characterized in that, The state decoupling module is used to receive the on-site working condition data, calculate and output the equivalent wear correlation index of the forming mold, specifically for: Real-time monitoring of angular acceleration, jamming discrete signal and emergency stop status signal in the field working condition data to determine the current abnormal state of external interference of the system; When the external interference is determined to be an abnormal state, the decoupling calculation is stopped, and the equivalent wear correlation index of the forming mold is maintained at the historical calculation value of the previous stable operating cycle.

6. The intelligent control system for processing parameters in a copper busbar manufacturing process according to claim 1, characterized in that, The feedforward control module is used to combine the comprehensive cooling intensity index, the material coordinate mapping table, the forming die equivalent wear correlation index, and the global control gain parameter to output the processing control command with physical constraints to the actuator, specifically for: Based on the meter counter feedback value in the on-site working condition data, locate the physical slice currently entering the forming area in the material coordinate mapping table, and extract the comprehensive cooling intensity index corresponding to the physical slice; Combining the comprehensive cooling intensity index extracted from the positioning, the current equivalent wear correlation index of the forming mold, and the first traction force compensation gain coefficient and the second traction force compensation gain coefficient in the global control gain parameters, the initial feedforward traction force command at the current sampling time is calculated using the feedforward traction force calculation formula.

7. The intelligent control system for processing parameters in a copper busbar manufacturing process according to claim 6, characterized in that, After calculating the initial feedforward traction force command, the feedforward control module outputs the processing control command with physical constraints, specifically for: The initial feedforward traction force command is compared with the traction force safety upper limit threshold. When the initial feedforward traction force command is greater than or equal to the traction force safety upper limit threshold, the traction force protection mechanism is triggered to limit the actual output traction force command to the traction force safety upper limit threshold and generate the target feedforward traction force command. Based on the difference between the initial feedforward traction force command and the traction force safety upper limit threshold, the cooperative deceleration compensation amount at the current sampling moment is calculated using the cooperative deceleration calculation formula. The coordinated deceleration compensation is superimposed on the main drive control loop and combined with the target feedforward traction command to generate the machining process control command with physical constraints.

8. The intelligent control system for processing parameters in a copper busbar manufacturing process according to claim 1, characterized in that, The processing control commands output by the feedforward control module include annealing compensation commands, and the processing control commands with physical constraints are output to the actuator, specifically for: The number of delay cycles required for the wire to cross the physical distance between the forming zone and the annealing heating zone is calculated by combining the traction displacement calculation. The annealing compensation instruction included in the processing control instruction is written into the first-in-first-out queue, and the read pointer offset of the underlying queue is continuously monitored. When the read pointer offset reaches the number of delay cycles, the corresponding annealing compensation instruction is removed from the first-in-first-out queue and sent to the execution mechanism.

9. The intelligent control system for processing parameters in a copper busbar manufacturing process according to claim 1, characterized in that, The parameter optimization module is used to perform abnormal sample rejection and preprocessing based on the end-product quality inspection data corresponding to the processing control instructions, specifically for: The quality inspection data of the terminal products are subjected to abnormal sample removal and preprocessing, and valid samples are retained; The feedforward control gain parameter update amount is calculated using the parameter incremental optimization formula based on the deviation between the actual size and the target size corresponding to the effective sample. The updated amount of the feedforward control gain parameters is added to the gain coefficient of the previous production batch to obtain the preliminary updated gain parameters.

10. The intelligent control system for processing parameters in a copper busbar manufacturing process according to claim 9, characterized in that, After obtaining the initially updated gain parameters, the parameter optimization module performs limit boundary checks and safety verification, and then outputs the updated global control gain parameters, specifically for: Perform limit boundary checks on the initially updated gain parameters. If the values ​​exceed the set boundaries, force the initially updated gain parameters to be clamped to the corresponding boundary values. Based on the boundary values ​​after forced clamping beyond the limit or the initially updated gain parameters before exceeding the limit, a safety review prompt is triggered on the human-computer interaction interface. After the safety review prompt is approved by the process personnel, the data is packaged and the updated global control gain parameters are output.

Images