A multi-station coordinated control and energy optimization method for hose printing

Abstract

This invention discloses a multi-station collaborative control and energy optimization method for tube printing, belonging to the field of industrial automation control technology. The method includes: constructing an event-triggered distributed collaborative control network; each station generating and reporting instantaneous time deviations when a physical completion event is triggered; a central collaborative optimizer solving for the optimal speed adjustment sequence based on model predictive control; establishing an energy optimization objective function based on the optimal speed sequence and ink layer thickness data using a long short-term memory network curing degree estimation model, and solving for the optimal power sequence under the premise of satisfying curing degree constraints and temperature constraints; the UV curing station executing the optimal power command and performing rolling optimization driven by events. This invention solves the technical problems of poor multi-station collaboration and low energy utilization in existing tube printing production lines, achieving global energy consumption minimization while ensuring printing registration accuracy and curing quality.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control technology, and in particular to a multi-station collaborative control and energy optimization method for hose printing. Background Technology

[0002] The tube printing process typically involves multiple continuous stations, including unwinding, multi-color printing, UV curing, and rewinding. The timing synchronization accuracy between these stations directly affects the printing registration quality, while the energy consumption of the UV curing unit accounts for a significant portion of the total production energy consumption. How to reduce curing energy consumption while ensuring printing accuracy has become a pressing technical challenge in this field.

[0003] Existing technologies have many shortcomings in handling multi-station collaboration and energy optimization problems. CN119758917A discloses a multi-station collaborative control platform for an automated production line of automotive sunroofs, which adopts a centralized quality assessment and post-event adjustment strategy, resulting in a lag in response to real-time timing deviations and lacking energy optimization capabilities. CN120370863A discloses a multi-parameter collaborative control system for PVC steel wire hose production, but it targets the extrusion molding process, which has different parameter coupling relationships with the printing and curing process, and it does not establish a real-time feedback model for curing effect and energy input. CN121277112A discloses a multi-station collaborative control method based on event triggering and look-ahead synchronization, but it is limited to the synchronous control of mechanical motion and does not incorporate energy optimization into the collaborative framework. CN121433090A discloses a multi-station collaborative control method using distributed PLC, but its cycle time adaptive algorithm does not involve the power collaborative control of the curing unit and lacks online energy optimization capabilities. CN121515603A discloses an intelligent collaborative control system for the UV curing process of a rotary printing press, but it is designed for paper-based substrates. Its temperature constraint model cannot be directly applied to flexible tubes, and no curing degree estimation model has been established, resulting in a control blind spot.

[0004] Therefore, there is an urgent need for a control scheme that can simultaneously solve the problems of multi-station timing coordination and energy dynamic optimization in tube printing, so as to minimize global energy consumption while ensuring printing registration accuracy and curing quality. Summary of the Invention

[0005] To address the technical problems existing in the prior art, this invention provides a multi-station collaborative control and energy optimization method for tube printing, comprising the following steps: S101. Construct a distributed collaborative control network and initialize the system. The central collaborative optimizer broadcasts synchronization signals to the local controllers of each workstation through real-time industrial Ethernet, performs clock synchronization based on the IEEE 1588 precision time protocol, and loads pre-configured production line parameters to put the system into a ready-to-run state. S102. Event Triggering and Station-Level Time Deviation Acquisition: When the local controller of the printing station detects a physical completion event, it records the actual completion time through a high-speed capture unit, calculates the instantaneous time deviation relative to the preset nominal completion time, and reports it. The central collaborative optimizer receives the instantaneous time deviation and generates a cumulative time deviation characterizing the cumulative registration error of the printed pattern through accumulation calculation. S103. Prospective alignment speed co-optimization based on model predictive control: The central co-optimizer reads the cumulative time deviation, and the model predictive controller embedded within it optimizes the cumulative alignment error squared sum and the speed adjustment change rate squared sum by minimizing the weighted sum of the cumulative alignment error squared sum and the speed adjustment change rate squared sum in the prediction time domain. Under the constraints of speed adjustment boundary and cumulative error terminal zeroing, it generates the optimal speed adjustment sequence for multiple future cycles through quadratic programming and issues the first speed adjustment command. S104. Dynamic Co-optimization of Energy and Speed: The central co-optimizer starts the energy optimization module in parallel. The energy optimization module constructs a dynamic energy optimization model in the prediction time domain based on the curing degree estimation model pre-trained by the long short-term memory network. The dynamic energy optimization model aims to minimize the total energy consumption of the UV curing unit. Under the constraints of the curing degree estimation value reaching the target value, the peak temperature not exceeding the threshold, and the power adjustable range, the optimal UV lamp power sequence is generated by solving the sequential quadratic programming algorithm, and the first optimal power command is issued. S105. Rolling time domain execution and feedback correction: Each station executes the speed adjustment command and the optimal power command. After one cycle is completed, the actual curing effect characterization value is collected and compared with the target curing degree value to generate an execution deviation. If the execution deviation exceeds the limit, the output of the energy optimization model is compensated and corrected in the next optimization cycle. S106. Energy pre-compensation processing for abnormal operating conditions: The central collaborative optimizer continuously monitors the optimal speed adjustment sequence. When it detects that the absolute value of the speed adjustment in the next cycle exceeds the stability threshold based on historical data statistics, it immediately triggers the energy pre-compensation mechanism. Based on the current speed fluctuation direction and amplitude, it calls the fuzzy rule base to generate a pre-compensation power instruction to cover or precede the regular power instruction.

[0006] As a preferred embodiment of this application, in step S101, the production line parameters include the nominal motion curve of each printing station, the nominal running time, the sensor trigger threshold of the physical completion event, the maximum and minimum power of the UV curing unit, and the safe temperature threshold.

[0007] As a preferred embodiment of this application, in step S102, the detection of the physical completion event refers to: the high-frequency response fiber optic sensor installed at the printing roller imprinting point position detecting the hardware interrupt signal generated when the roller reaches the imprinting point; the instantaneous time deviation and the workstation identifier are encapsulated into a standard Ethernet data frame and then published through the industrial real-time network.

[0008] As a preferred embodiment of this application, in step S103, the weight coefficients in the optimization objective function can be adjusted online through the human-computer interaction module to adapt to the different requirements of the production line for printing accuracy and stability.

[0009] As a preferred embodiment of this application, in step S104, the input feature vector of the curing degree estimation model includes: the UV lamp power setting value of the current cycle, the real-time temperature of the curing area on the surface of the hose material measured by an infrared thermometer, the printing speed, and the real-time ink layer thickness fed back by the online visual inspection system.

[0010] As a preferred embodiment of this application, in step S105, the acquisition of the actual curing effect characterization value includes: acquiring real-time data through a temperature sensor and a UV light intensity sensor of the curing area, and obtaining online indirect measurement values ​​of the gloss or adhesion of the current product through an online visual inspection system.

[0011] As a preferred embodiment of this application, in step S106, the fuzzy rule base is pre-generated, and its rules include: if the speed will decrease by a certain value, the UV lamp power will be reduced by a corresponding proportion in advance; if the speed will increase, the power will be increased in advance.

[0012] As a preferred embodiment of this application, the method further includes: performing registration error detection and ink layer thickness detection through an online visual inspection system deployed after the printing station and before the UV curing station, and reporting the calculated cumulative registration error value and ink layer thickness data to the central collaborative optimizer as feedback input for steps S102 and S104.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: 1) Achieving a forward-looking improvement in printing registration accuracy based on event-triggered and model predictive control (MPC): This invention solves the problem of lag in traditional control response by using a high-frequency response fiber optic sensor to detect physical completion events and trigger hardware interrupts in S102, and then using a model predictive controller (MPC) to solve for the optimal speed sequence in S103. Specifically, the system no longer relies solely on the current instantaneous deviation for adjustment, but instead records the deviation between the actual and nominal times through a high-speed acquisition unit to generate cumulative registration error. In the prediction time domain, the MPC algorithm comprehensively considers the error accumulation trend and the rate of change of speed adjustment over multiple future cycles, and solves for an optimal speed adjustment sequence through quadratic programming. This forward-looking control strategy can smoothly eliminate cumulative errors and avoid the oscillations or lags caused by simple feedback control, thereby achieving a significant improvement in printing registration accuracy under physical constraints.

[0014] 2) Minimizing UV Curing Energy Consumption Based on LSTM Curing Degree Model and Dynamic Co-optimization: This invention solves the problem of mismatch between curing energy consumption and production speed in existing technologies by constructing a curing degree estimation model based on a Long Short-Term Memory (LSTM) network in S104 and combining it with a sequential quadratic programming algorithm. Specifically, the energy optimization module starts in parallel, using a pre-trained LSTM network to accurately estimate the current curing degree using real-time ink layer thickness, printing speed, temperature, and power as input feature vectors. Based on this, a dynamic optimization model is constructed with the goal of minimizing total energy consumption. This model dynamically calculates the optimal UV lamp power sequence under the premise of meeting curing degree requirements and temperature safety constraints. Compared with traditional fixed power or coarse control that only follows speed, this scheme achieves precise matching between energy input and printing process parameters (such as ink layer thickness variation). While ensuring curing quality, it avoids energy waste caused by excessive irradiation and effectively reduces the global energy consumption of the UV curing unit.

[0015] 3) Introducing an energy pre-compensation mechanism for abnormal operating conditions to enhance system robustness: This invention solves the problem of easily impacted curing quality under large speed fluctuations by setting speed fluctuation monitoring and fuzzy rule base pre-compensation in S106. Specifically, the central co-optimizer continuously monitors the speed adjustment sequence generated in S103. Once it detects that the absolute value of the speed adjustment in the next cycle exceeds the stability threshold based on historical data statistics (i.e., a large speed fluctuation is predicted), the system immediately triggers the energy pre-compensation mechanism. This mechanism does not wait for the regular optimization cycle but directly calls the fuzzy rule base (e.g., reducing power in advance when the speed decreases, and increasing power in advance when the speed increases) to generate pre-compensation power instructions that cover regular instructions. This proactive pre-compensation strategy ensures that changes in UV power precede changes in speed, effectively offsetting the impact of speed disturbances on curing energy accumulation, thus maintaining curing stability under abnormal or drastic speed change conditions and enhancing system robustness.

[0016] 4) Eliminating control blind spots and establishing a quantitative closed loop for quality and energy consumption: This invention, through rolling time-domain execution and feedback correction in S105, solves the defect of existing technologies that cannot directly optimize based on curing effect. Specifically, the system uses an online visual inspection system to obtain characterization values ​​such as gloss or adhesion, and compares them with the target curing degree to generate execution deviation. If the deviation exceeds the limit, the system will compensate and correct the output of the energy optimization model in the next cycle. At the same time, an LSTM network is used to process nonlinear dynamic relationships, establishing a precise mapping between power, speed, ink layer thickness, and curing degree. This transforms the control objective from simple "temperature or power control" to direct "curing quality control," eliminating control blind spots and achieving global energy consumption minimization while ensuring product quality. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a general flowchart of the multi-station collaborative control and energy optimization method provided in the embodiments of the present invention.

[0020] Figure 2 This is a flowchart of event triggering and workstation-level time deviation acquisition provided in an embodiment of the present invention.

[0021] Figure 3 This is a flowchart of the forward-looking registration-speed collaborative optimization provided in the embodiments of the present invention.

[0022] Figure 4 This is a flowchart of the energy-velocity dynamic collaborative optimization provided in the embodiments of the present invention.

[0023] Figure 5 This is an architecture diagram of the multi-station collaborative control and energy optimization system provided in the embodiments of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0025] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0026] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" and "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0027] Example 1 This invention provides a multi-station collaborative control and energy optimization method for tube printing. This method captures real-time deviations at each printing station through an event-triggered mechanism, proactively adjusts the printing speed using model predictive control to eliminate accumulated registration errors, and dynamically optimizes the power setting of the UV curing unit based on a cure degree estimation model, thereby minimizing global energy consumption while ensuring printing and curing quality. Figures 1 to 4 As shown, the specific steps include: S101. Constructing a distributed collaborative control network and system initialization. First, the central coordinating optimizer broadcasts synchronization signals via real-time industrial Ethernet to local controllers deployed at the unwinding station, first printing station, second printing station, UV curing station, and rewinding station. It then performs clock synchronization based on the IEEE 1588 precision time protocol, ensuring that the local clock deviation of all control nodes across the network is locked within ±1 microsecond, guaranteeing a unified time base for subsequent event timestamps. Simultaneously, the central coordinating optimizer loads pre-configured production line parameters, including the nominal motion curves, nominal runtime, sensor trigger thresholds for physically completed events, maximum and minimum power of the UV curing unit, and safe temperature thresholds for each printing station. After initialization, the system enters a pre-triggered operation state, awaiting the start of the first production cycle.

[0028] S102, Event Triggering and Workstation-Level Time Deviation Acquisition When the first printing station begins a production cycle, its local controller drives the servo unit to run according to the nominal motion curve. A high-frequency response fiber optic sensor installed at the printing roller's imprinting point monitors the physical state in real time. Once it detects the physical completion event of the roller reaching the imprinting point, the sensor immediately generates a hardware interrupt signal. The local controller of the first printing station records the arrival time of this hardware interrupt signal through its built-in high-speed capture unit, recording it as the actual completion time. Simultaneously, the local controller retrieves the preset nominal completion time of the station from its local memory and generates an instantaneous time deviation through subtraction. This instantaneous time deviation, along with the station identifier, is encapsulated into a standard Ethernet data frame and published to the bus via the industrial real-time network.

[0029] The communication management unit of the central coordinating optimizer continuously monitors network data. Upon receiving the instantaneous time deviation from the first printing station, the data receiving unit parses the standard Ethernet data frame and passes it to the internal cumulative deviation calculator. The cumulative deviation calculator performs an accumulation operation to obtain the cumulative time deviation up to the current station. For the first station, the cumulative time deviation is equal to the instantaneous time deviation. When an event is triggered at a subsequent station (such as the second printing station), the local controller of that station also calculates its own instantaneous deviation relative to the nominal reference time. The central coordinating optimizer then executes the accumulation logic, adding the instantaneous deviation of the current station to the cumulative deviation transmitted from the previous station, forming a deviation accumulation chain that propagates along the station sequence. This cumulative time deviation directly represents the cumulative registration error of the printed pattern on the flexible tube material. This process ensures that the system can perceive and quantify the deviation between the actual timing and the nominal timing of each printing cycle in real time.

[0030] S103, Forward-looking registration-velocity co-optimization based on model predictive control The parameter calculation unit of the central co-optimizer reads the cumulative time deviation (i.e., cumulative registration error) calculated in S102. The parameter calculation unit embeds a model predictive controller, which first constructs a prediction time domain of length H, covering the next H printing cycles after the current workstation. The goal of the model predictive controller is to solve for the optimal speed adjustment for a set of future H cycles, ensuring that the cumulative registration error can be smoothly eliminated while satisfying physical constraints.

[0031] Specifically, the parameter calculation unit first defines the state variable as the cumulative registration error at the end of the k-th prediction cycle, and defines the control input as the speed adjustment amount (i.e., the rate of change relative to the nominal speed) of the k-th cycle. Based on kinematic relationships, a discrete state-space model is established, which describes how the speed adjustment amount affects the evolution of the cumulative error. Then, an optimization objective function is constructed, consisting of a weighted sum of two terms: the first term penalizes the sum of squares of the cumulative registration error of each cycle in the prediction time domain to ensure printing accuracy; the second term penalizes the sum of squares of the rates of change of speed adjustments in adjacent cycles to limit the mechanical shock caused by excessive acceleration. The weight coefficients in the objective function are adjusted online through the human-machine interaction module according to the different accuracy and stability requirements of the production line.

[0032] Based on the objective function, the parameter calculation unit sets constraints, including upper and lower limits for speed adjustment, determined by the physical limits of the servo motor. It also sets a cumulative error terminal constraint, requiring the cumulative time deviation to reach zero at the end of the Hth cycle, ensuring the deviation is completely absorbed within the prediction time domain. Finally, the parameter calculation unit calls the embedded quadratic programming solver to perform a rolling solution to the optimization problem, generating a sequence of optimal speed adjustments. Based on the first value in the sequence, the instruction generation unit generates the speed synchronization control instruction for the next cycle of the first printing station and sends it to the local controller of the first printing station for execution via the communication management unit. Simultaneously, the central co-optimizer temporarily stores the entire optimal speed sequence in memory as input for the next energy optimization step.

[0033] S104, Energy-Velocity Dynamic Co-optimization While S103 is being executed, the parallel computing thread of the central co-optimizer starts the energy optimization module. The energy optimization module first obtains real-time ink layer thickness data from the online visual inspection system via the data acquisition interface. Simultaneously, it reads the optimal speed sequence calculated and temporarily stored in S103.

[0034] The core of the energy optimization module is a curing degree estimation model pre-trained using a Long Short-Term Memory (LSTM) network. The input layer of the curing degree estimation model receives the following feature vectors: the UV lamp power setting for the current cycle, the real-time temperature of the cured area on the hose material surface measured by an infrared thermometer, the printing speed, and the ink layer thickness. The output layer of the curing degree estimation model outputs a scalar value, namely the estimated curing degree of the hose surface. The curing degree estimation model has been trained using historical production data and can accurately map the nonlinear dynamic relationship between input parameters and curing effect. It contains multiple memory units and gating structures to capture long-term dependencies over time.

[0035] Based on the curing degree estimation model, the energy optimization module constructs a dynamic energy optimization model. The goal of the dynamic energy optimization model is to minimize the total energy consumption of the UV curing unit within the prediction time domain, finding a set of UV lamp power settings for the next H cycles that minimizes the total energy consumption while meeting curing degree requirements and temperature constraints. Specifically, the objective function is to minimize the sum of the products of power and time for each cycle; the curing degree constraint requires that the estimated curing degree of the hose surface for each cycle must reach the target value required by the process; the temperature constraint requires that the peak temperature of the curing region must not exceed the material's tolerance limit; and the power physical constraint limits the adjustable range of the UV lamp.

[0036] The parameter calculation unit uses a sequential quadratic programming algorithm to solve this nonlinear constraint optimization problem. The solver generates a set of optimal power sequences through iterative calculation. The instruction generation unit extracts the first optimal power value, encapsulates it into a UV light source power control instruction, and sends it to the local controller of the UV curing station for execution via the communication management unit. This process enables dynamic tracking of energy input to changes in printing speed, avoiding energy waste or insufficient curing caused by traditional fixed power settings.

[0037] S105, Rolling Time Domain Execution and Feedback Correction The local controller at the UV curing station receives the optimal power command and adjusts the output power of the UV light source through its internal power regulation unit. After one cycle is completed, the temperature sensor and UV illuminance sensor in the curing area re-collect real-time data and obtain the actual curing effect characterization values ​​of the current product, such as online indirect measurements of gloss or adhesion, through an online visual inspection system.

[0038] The data receiving unit of the central co-optimizer collects this new feedback data and executes deviation judgment logic: it compares the actual curing effect value with the target curing degree value required by the process and calculates the execution deviation. If the execution deviation exceeds the preset allowable deviation range, the central co-optimizer triggers compensation adjustment logic, and in the next optimization cycle, it compensates and corrects the energy optimization model output by adjusting the weights or introducing a deviation penalty term.

[0039] Simultaneously, the system uses the occurrence of the physical completion event of the second printing station as a new trigger signal to repeatedly execute S102 to S105. The entire control process is event-driven, using a fixed-length prediction time domain as a window, and progresses forward in a rolling manner, achieving real-time dynamic optimization of the continuous production process.

[0040] S106, Energy pre-compensation processing for abnormal operating conditions The central co-optimizer continuously monitors the speed adjustment sequence obtained in S103. When it detects that the absolute value of the speed adjustment in the next cycle exceeds the stability threshold obtained by analyzing historical speed data using statistical process control methods, it determines that a large speed fluctuation is about to occur. At this point, the energy optimization module does not wait for the next optimization cycle but immediately triggers the energy pre-compensation mechanism.

[0041] The energy pre-compensation mechanism performs the following actions: based on the current direction and magnitude of speed fluctuations, it invokes a pre-generated fuzzy rule base. For example, if the speed is expected to decrease by a certain value, the UV lamp power is reduced by the corresponding proportion in advance to prevent instantaneous over-curing caused by the speed decrease; if the speed is expected to increase, the power is increased in advance to make up for the energy gap. This pre-compensation instruction is immediately inserted into the execution queue, overriding or preceding the regular power instruction, thereby effectively suppressing the impact of speed disturbances on curing quality. This pre-compensation logic is executed in the parameter calculation unit of the central co-optimizer, ensuring that the pre-compensation decision is completed before the speed control instruction is issued.

[0042] Example 2 This invention also provides a multi-station collaborative control and energy optimization system for flexible tube printing, wherein the components are connected via a PROFINET IRT real-time industrial Ethernet bus to achieve high-speed, deterministic data transmission. Figure 5 As shown, it specifically includes: Central Coordinator The central co-optimizer is the core decision-making unit of the entire system. Its physical carrier is an industrial control computer with powerful computing capabilities (such as a high-performance multi-core CPU and an optional AI accelerator card) to meet the real-time computing needs of model predictive control, long short-term memory network inference, and sequential quadratic programming solutions. The central co-optimizer integrates a data receiving unit, a parameter calculation unit, an instruction generation unit, and a communication management unit.

[0043] The data receiving unit consists of a network interface controller and a direct memory access driver. It captures and parses all event data frames from the local controller, as well as image processing result data packets from the online vision inspection system, through a bus listening mechanism. The parsed data is formatted into a unified internal data structure and stored in a real-time database for use by other units.

[0044] The parameter calculation unit is the "brain" of the system, internally deploying three core algorithm engines: a model predictive controller, a long short-term memory network solidification estimator, and a sequence quadratic programming optimization solver. In addition, this unit includes a pre-compensation logic module with an embedded fuzzy rule base for performing energy pre-compensation processing. The model predictive controller receives the accumulated time deviation, performs quadratic programming optimization, and outputs the optimal velocity sequence; the solidification estimator is a trained neural network model deployed on a deep learning inference engine, receiving real-time data such as ink layer thickness and velocity, performing rapid forward computation, and outputting a solidification estimate; the optimization solver receives the velocity sequence and solidification constraints, performs constraint optimization, and outputs the optimal power sequence.

[0045] The instruction generation unit converts the mathematical solution output by the parameter calculation unit into a standard control instruction frame conforming to the industrial Ethernet protocol. The instruction frame contains the target address, opcode, and operands, ensuring that each instruction can be accurately recognized and executed by the corresponding local controller.

[0046] The communication management unit is responsible for managing the real-time performance of bus communication, including packet priority scheduling, periodic and aperiodic data transmission management, and network diagnostics. It ensures that optimization commands are delivered to the target controller within a deterministic timeframe (e.g., a 1ms period), while simultaneously monitoring network status in real time and quickly switching to redundant paths in the event of link interruptions or node failures, thus guaranteeing system reliability.

[0047] Local controller Local controllers are deployed at the unwinding station, the first printing station, the second printing station, the UV curing station, and the rewinding station. Each controller is based on an industrial programmable logic controller or an embedded driver with hardware interrupt handling capabilities. Each local controller integrates an event capture and deviation calculation module, a motion control execution module, or an energy regulation execution module.

[0048] For the local controller at the printing station, the event capture and deviation calculation module includes a high-speed digital input channel connected to the physical completion event sensor. When the sensor signal is triggered, the hardware immediately captures the current free-running clock count, generates the actual completion time, and the deviation calculation logic in the firmware runs accordingly. It then calculates the instantaneous time deviation through subtraction and stores it in the transmission buffer for later reporting. The motion control execution module receives speed commands from the central co-optimizer. The internal servo driver generates a pulse-width modulation waveform based on the target speed curve using a vector control algorithm, precisely controlling the torque, speed, and position of the servo motor.

[0049] For the local controller of the UV curing station, the energy regulation execution module includes a power regulation unit. The power regulation unit uses phase-shift pulse width modulation (PWM) or digital-to-analog conversion (DAC) technology to convert digital power commands into 0-10V analog signals or PWM waves, controlling the duty cycle of the UV power supply's switching transistors to continuously adjust the output power of the UV-LED lamp array. The light source status monitoring unit samples the voltage, current, and temperature of the lamps in real time, and feeds this data back to the central co-optimizer after analog-to-digital conversion for status monitoring and fault early warning.

[0050] Online visual inspection system The online visual inspection system is deployed after the printing station and before the UV curing station, and includes a high-speed industrial camera and an embedded image processing unit. The high-speed industrial camera is triggered by an encoder signal to take a picture as the printed material passes by, capturing images of the printed pattern and pre-reserved registration marks. The embedded image processing unit runs image processing algorithms. First, it performs registration error detection, calculating the positional deviation between the current printed pattern and the standard template using template matching or edge detection algorithms. This deviation is converted into a cumulative registration error value, serving as feedback input for model predictive control. Next, it performs ink layer thickness detection, calculating the ink layer thickness data through optical reflectivity analysis at a specific wavelength or by measuring the ink height using a 3D laser profilometer. This data is then packaged together with the image data and reported to the central co-optimizer via Ethernet.

[0051] Human-computer interaction module The human-machine interface (HMI) module is an industrial touchscreen running monitoring and data acquisition software. It communicates with the central co-optimizer via the OPC unified architecture protocol. The HMI module's functions include parameter configuration, status visualization, and alarm management. Operators can input parameters such as nominal cycle time, weighting coefficients, and temperature thresholds via the touchscreen, and these parameters are written to the central co-optimizer's configuration register. Simultaneously, the HMI module reads real-time data from the central co-optimizer, including real-time speed curves, UV power curves, cumulative deviation curves, and energy consumption statistics, and dynamically displays these in chart form on the screen. When the system experiences communication interruptions, sensor failures, or insufficient curing, the HMI module immediately displays alarm information and logs the information.

[0052] Through the precise coordination and data interaction of the above components, this system achieves full-link collaborative control and optimization in the tube printing process, from deviation perception and speed planning to dynamic energy matching.

[0053] Example 3 The present invention also provides an electronic device, including: a processor, a transmitting device, an input device, an output device, and a memory. The processor may be implemented using a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit, or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory may be implemented using a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM), and is used to store computer program code. The computer program code includes computer instructions. When the processor executes the computer instructions, the electronic device executes a method as described in any of the above possible implementation methods.

[0054] Example 4 The present invention also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor of an electronic device, cause the processor to perform a method as described in any of the above possible implementations.

[0055] In the description of this specification, the references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0056] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A multi-station collaborative control and energy optimization method for flexible tube printing, characterized in that, Includes the following steps: S101. Construct a distributed collaborative control network and initialize the system through a central collaborative optimizer; S102, Receive physical completion events reported by the local controllers of each workstation; Based on the physical completion event, the instantaneous time deviation of each workstation relative to the nominal time is calculated, and the cumulative time deviation is generated by accumulation. S103. Collaborative optimization of alignment speed based on model predictive control: The central collaborative optimizer reads the cumulative time deviation, and the model predictive controller embedded within it optimizes the cumulative alignment error square sum and the speed adjustment change rate square sum by minimizing the weighted sum of the cumulative alignment error square sum and the speed adjustment change rate square sum in the prediction time domain. Under the constraints of speed adjustment boundary and cumulative error terminal zeroing, it generates the optimal speed adjustment sequence for multiple future cycles through quadratic programming and issues the first speed adjustment command. S104. Dynamic Co-optimization of Energy and Speed: The central co-optimizer starts the energy optimization module in parallel. The energy optimization module constructs a dynamic energy optimization model in the prediction time domain based on the curing degree estimation model pre-trained by the long short-term memory network. The dynamic energy optimization model aims to minimize the total energy consumption of the UV curing unit. Under the constraints of the curing degree estimation value reaching the target value, the peak temperature not exceeding the threshold, and the power adjustable range, the optimal UV lamp power sequence is generated by solving the sequential quadratic programming algorithm, and the first optimal power command is issued. S105. Rolling time-domain execution and feedback correction: Each station executes the speed adjustment command and the optimal power command. After one cycle is completed, the actual curing effect characterization value is collected and compared with the target curing degree value to generate an execution deviation. If the execution deviation exceeds the limit, the output of the energy optimization model is compensated and corrected in the next optimization cycle. S106. Energy pre-compensation processing for abnormal operating conditions: The central collaborative optimizer continuously monitors the optimal speed adjustment sequence. When it detects that the absolute value of the speed adjustment in the next cycle exceeds the stability threshold based on historical data statistics, it immediately triggers the energy pre-compensation mechanism. Based on the current speed fluctuation direction and amplitude, it calls the fuzzy rule base to generate a pre-compensation power instruction to cover or precede the regular power instruction.

2. The method according to claim 1, characterized in that, In step S101, the central collaborative optimizer broadcasts a synchronization signal to the local controller of each workstation via real-time industrial Ethernet, performs clock synchronization based on the IEEE 1588 precision time protocol, and loads pre-configured production line parameters, so that the system enters the ready-to-run state.

3. The method according to claim 2, characterized in that, In step S101, the production line parameters include the nominal motion curve of each printing station, the nominal running time, the sensor trigger threshold of the physical completion event, the maximum and minimum power of the UV curing unit, and the safe temperature threshold.

4. The method according to claim 3, characterized in that, In step S102, the physical completion event includes: a hardware interrupt signal generated when a high-frequency response fiber optic sensor installed at the printing roller imprinting point detects that the roller has reached the imprinting point; the instantaneous time deviation and the workstation identifier are encapsulated into a standard Ethernet data frame and then published through an industrial real-time network.

5. The method according to claim 4, characterized in that, In step S103, the weight coefficients in the optimization objective function can be adjusted online through the human-computer interaction module to adapt to the different requirements of the production line for printing accuracy and stability.

6. The method according to claim 5, characterized in that, In step S104, the input feature vector of the curing degree estimation model includes: the UV lamp power setting value of the current cycle, the real-time temperature of the curing area on the surface of the hose material measured by an infrared thermometer, the printing speed, and the real-time ink layer thickness fed back by the online visual inspection system.

7. The method according to claim 6, characterized in that, In step S105, the acquisition of actual curing effect characterization values ​​includes: acquiring real-time data through temperature sensors and UV light intensity sensors in the curing area, and obtaining online indirect measurement values ​​of gloss or adhesion of the current product through an online visual inspection system.

8. The method according to any one of claims 1-7, characterized in that, In step S106, the fuzzy rule base is pre-generated, and its rules include: if the speed will decrease by a certain value, the UV lamp power will be reduced by a corresponding proportion in advance; if the speed will increase, the power will be increased in advance.

9. The method according to claim 1, characterized in that, The method further includes: using an online visual inspection system deployed after the printing station and before the UV curing station to perform registration error detection and ink layer thickness detection, and reporting the calculated cumulative registration error value and ink layer thickness data to the central collaborative optimizer as feedback input for steps S102 and S104.

Images