Multi-axis servo linkage and pneumatic-electric coordinated control method for fully automatic screen printing machine

Abstract

The present invention provides a multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic silk screen printing machine, which relates to the field of data processing technology. The method comprises: obtaining trajectory data of a silk screen printing platform; screening clamping response sample data with similar trajectories in a printing cycle from a historical database, constructing a clamping response prediction model, and predicting the predicted response time of a pneumatic clamp; calculating the pre-trigger timing of a pneumatic clamp pre-trigger instruction, and then calculating a predicted threshold value of the distance between the silk screen printing platform and a target position; when the distance between the servo axis and the target position is less than the predicted threshold value, outputting a pre-trigger instruction, collecting state data during the response process of the pneumatic clamp, and calculating the actual response time of the pneumatic clamp based on the data; comparing the actual response time with the predicted response time, and automatically adjusting the pre-trigger timing of the pneumatic clamp for the next printing cycle; the present invention improves the intelligence and accuracy of multi-axis servo linkage and pneumatic-electrical coordinated control.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine. Background Art

[0002] Existing fully automatic screen printing technology typically uses a multi-axis servo motor as the core drive unit. A motion control system provides coordinated control of multiple motion units, such as the screen printing platform, squeegee, and screen frame, to achieve functions such as position adjustment and printing rhythm matching during the printing process. Furthermore, auxiliary actions (such as clamping, lifting, and embossing) are often performed by pneumatic components. Conventional practice is to control the servo drive and pneumatic components separately through a PLC program, with each operating according to a pre-set program. In actual operation, this independent control method can generally meet the automation requirements of screen printing operations.

[0003] However, when performing multi-layer overprinting of complex patterns, the separation of servo and pneumatic control in existing technologies may lead to uncoordinated system responses. Taking high-precision circuit board silk screen printing as an example, when the servo-driven silk screen printing platform completes positioning, the pneumatic clamp must immediately clamp the workpiece. If the clamping action lags due to signal delays or synchronization errors between the servo control and pneumatic control at this time, it may cause micro-displacement of the workpiece, thereby affecting the overprinting accuracy. This defect is particularly evident in the production of high-precision electronic components, which may cause printing deviations or even product scrapping, making it difficult to meet the position synchronization requirements of high-standard manufacturing processes. Summary of the Invention

[0004] The purpose of the present invention is to provide a multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine, aiming to solve the problems mentioned in the background technology.

[0005] In order to solve the above technical problems, the technical solutions of the present invention are as follows:

[0006] A multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine, the method comprising:

[0007] Obtain the positioning parameters of the target workpiece, and use them to drive the screen printing platform to move, collect and record the motion trajectory of the screen printing platform in real time, and obtain trajectory data;

[0008] Based on the trajectory data, the clamping response sample data with similar trajectories in the printing cycle are screened from the historical database, and a clamping response prediction model is constructed to predict the predicted response time of the pneumatic clamp under the current printing cycle;

[0009] Based on the predicted response time and trajectory data, the pre-trigger timing for outputting the pneumatic fixture pre-trigger command in advance is calculated, and then the predicted threshold of the screen printing platform's distance from the target position is calculated;

[0010] When the distance the servo axis moves to the target position is less than the prediction threshold, a pre-trigger instruction for the pneumatic fixture is output, the status data of the pneumatic fixture during the response process is collected, and the actual response time of the pneumatic fixture is calculated based on the status data;

[0011] The actual response time of the pneumatic clamp is compared with the predicted response time to obtain a response difference value, and the pre-trigger timing of the pneumatic clamp for the next printing cycle is automatically adjusted according to the response difference value.

[0012] Preferably, based on the trajectory data, clamping response sample data with similar trajectories in the printing cycle are screened from the historical database, and a clamping response prediction model is constructed to predict the predicted response time of the pneumatic clamp under the current printing cycle, including:

[0013] Based on the trajectory data, the clamping response sample data with similar trajectories in the printing cycle are screened from the historical database to construct a reference data set;

[0014] Cluster analysis was performed on the clamping response time recorded in the reference dataset to extract the correlation features between the screen printing platform speed, acceleration and the pneumatic clamp response time, and generate the response time feature vector.

[0015] A clamping response prediction model is constructed based on the response time feature vector and trajectory data;

[0016] The predicted response time of the pneumatic clamp under the current working conditions is output through the clamping response prediction model.

[0017] Preferably, the pre-trigger timing for outputting the pneumatic clamp pre-trigger instruction in advance is calculated based on the predicted response time and trajectory data, and then the predicted threshold of the screen printing platform from the target position is calculated, including:

[0018] Based on the trajectory data, the speed and acceleration parameters of the screen printing platform are calculated, and the remaining running time required for the screen printing platform to move from the current position to the target position is inferred.

[0019] Calculate the pre-triggering timing for issuing the pneumatic pre-triggering command in advance based on the remaining running time and predicted response time;

[0020] According to the pre-trigger timing, combined with the current speed and acceleration of the screen printing platform, the position offset corresponding to the time point is inferred as the prediction threshold of the screen printing platform from the target position.

[0021] Preferably, automatically adjusting the pre-triggering timing of the pneumatic clamp of the next printing cycle according to the difference value includes:

[0022] Based on the preset response tolerance threshold, calculate whether the response difference value is less than it;

[0023] When the response difference value is less than the response tolerance threshold, a proportional factor is used to perform a linear weighted correction on the response difference value to obtain a first compensation value;

[0024] When the response difference value is greater than the response tolerance threshold, a weighted average error is calculated based on the response difference values ​​recorded for a preset number of rounds, and the compensation amplitude is adjusted based on the weighted average error to obtain a second compensation value;

[0025] The first compensation value or the second compensation value is time-weightedly superimposed on the pre-trigger timing in the previous printing cycle to form a corrected pre-trigger timing, wherein the first compensation value or the second compensation value acts directly on the original trigger time as an offset.

[0026] Preferably, cluster analysis is performed on the clamping response times recorded in the reference data set to extract the correlation features between the screen printing platform speed, acceleration and the pneumatic clamp response time, and generate a response time feature vector, including:

[0027] Extract the screen printing platform speed and acceleration parameters corresponding to each clamping response time in the reference data set to form a parameter set;

[0028] Cluster the parameter set, classify samples with similar response characteristics into the same category, and obtain the sample set under each category;

[0029] For each category of sample set, the mean or median of the speed, acceleration, and response time of the category is calculated to form category feature data;

[0030] All category feature data are vectorized to form a response time feature vector containing speed, acceleration and response time statistics.

[0031] Preferably, a clamping response prediction model is constructed based on the response time feature vector and trajectory data, including:

[0032] The response time characteristic vector and the trajectory data collected during the current printing cycle are used as input parameters to determine the input variables;

[0033] Based on the relationship between historical input variables and their corresponding actual response time of the pneumatic clamp, the mapping relationship between the two is parameter fitted to establish a clamping response prediction model;

[0034] The prediction accuracy of the clamping response prediction model is verified by comparing it with the actual response time of the pneumatic clamp collected in the current or subsequent printing cycle. The model parameters are corrected according to the comparison results to obtain the corrected clamping response prediction model.

[0035] Preferably, the velocity and acceleration parameters of the screen printing platform are calculated based on the trajectory data, and the remaining running time required for the screen printing platform to move from the current position to the target position is inferred, including:

[0036] According to the trajectory data, the speed of the screen printing platform at the current time point is obtained by calculating the position change and time interval between adjacent sampling time points;

[0037] Calculate the velocity change and time interval of adjacent sampling time points to obtain the acceleration of the screen printing platform at the current time point;

[0038] Based on the position information, speed and acceleration parameters of the screen printing platform at each time point, the distance between the current position of the screen printing platform and the target position is determined. Based on the principles of kinematics, the remaining running time required for the screen printing platform to move to the target position at the current speed and acceleration is determined.

[0039] When it is detected that the acceleration of the screen printing platform is zero, the distance between the current position and the target position is divided by the current speed to obtain the remaining running time required to move to the target position at a constant speed.

[0040] Preferably, calculating the pre-triggering timing for issuing the pneumatic pre-triggering instruction in advance according to the remaining running time and the predicted response time includes:

[0041] By comparing the remaining running time with the predicted response time, it is determined that when the remaining running time of the screen printing platform is earlier than the predicted response time, a pre-trigger instruction of the pneumatic clamp is issued in advance;

[0042] Subtract the predicted response time from the remaining running time to calculate the specific value of the pre-trigger instruction and obtain the pre-trigger timing at which the screen printing platform should issue the pre-trigger instruction in advance;

[0043] When the pre-trigger timing is less than the preset minimum safety advance value, the minimum safety advance value is used as the pre-trigger timing for issuing the pre-trigger instruction, and the pre-trigger timing is output.

[0044] Preferably, when the response difference value is greater than the response tolerance threshold, a weighted average error is calculated based on the response difference values ​​recorded for a preset number of rounds, and the compensation amplitude is adjusted based on the weighted average error to obtain a second compensation value, including:

[0045] The response difference values ​​generated in the current printing cycle are combined with the response difference values ​​recorded in a preset number of rounds to form an error processing data set;

[0046] Based on the weight distribution principle set in descending order according to the distance from the current printing cycle, a weight coefficient is assigned to each response difference value in the error processing data set;

[0047] Perform weighted summation on all response difference values ​​that have been assigned weight coefficients, and divide the weighted summation value by the sum of the weight coefficients to obtain the weighted average error used for compensation adjustment;

[0048] According to the absolute value and change trend of the weighted average error, the compensation amplitude used to correct the pre-trigger timing of the next cycle is adjusted so that the compensation amplitude automatically adapts to the cumulative error trend to generate a second compensation value.

[0049] Preferably, performing time-weighted superposition of the first compensation value or the second compensation value and the pre-trigger timing in the previous printing cycle to form a corrected pre-trigger timing includes:

[0050] The first compensation value or the second compensation value in the current printing cycle is used as input data together with the pre-trigger timing recorded in the previous printing cycle;

[0051] Determining a weighting coefficient for the current printing cycle according to the first compensation value or the second compensation value in the current printing cycle, a historical error change trend, and a preset correction priority;

[0052] The first compensation value or the second compensation value is multiplied by a weighting coefficient to obtain a weighted compensation amount, and the weighted compensation amount is added to the pre-trigger timing in the previous printing cycle to obtain a corrected pre-trigger timing.

[0053] The above solution of the present invention includes at least the following beneficial effects:

[0054] The present invention breaks through the technical bottleneck of the existing technology in which the servo platform and the pneumatic clamp are independent of each other and lack high-precision synchronization by introducing adaptive compensation control of multi-axis servo linkage and pneumatic-electric coordination into the control process of the fully automatic screen printing machine. Compared with the traditional approach of using PLC to control the servo and pneumatic systems separately, the present invention can dynamically predict the response timing of the pneumatic clamp based on the real-time collected screen printing platform trajectory data and historical response samples, and accordingly calculate the optimal pre-trigger timing for the pneumatic clamp to start in advance before the platform moves to the target position. In this way, before the printing platform reaches the target position, the pneumatic clamp has been accurately activated in advance, so that the clamping action and platform positioning are highly coordinated, effectively eliminating the micro-displacement of the workpiece caused by signal delay and asynchrony.

[0055] The present invention further forms a feedback loop by collecting the actual response state of the fixture and continuously comparing it with the predicted response time, which can automatically compensate and correct the synchronization error in each printing cycle. The system automatically adjusts the fixture pre-trigger timing for the next cycle according to the size of the error, realizing the adaptive optimization of collaborative control parameters during multiple rounds of printing, ensuring the consistency of workpiece positioning and the qualified rate of finished products in high-precision multi-layer overprinting and high-standard electronic component production. This solution not only improves the adaptability of fully automatic screen printing machines to complex processes and high-precision manufacturing, but also reduces the risk of rework and scrap due to synchronization deviations, significantly improving the intelligence, stability and manufacturing quality of the production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a flowchart of the multi-axis servo linkage and gas-electric coordinated control method of the fully automatic screen printing machine provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0058] like Figure 1 As shown, an embodiment of the present invention proposes a multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine, the method comprising:

[0059] S100, obtaining positioning parameters of the target workpiece, and using the positioning parameters to drive the screen printing platform to move, collecting and recording the motion trajectory of the screen printing platform in real time to obtain trajectory data, wherein the trajectory data includes position information of the screen printing platform at different sampling time points and corresponding time information;

[0060] S200, based on the trajectory data, filtering clamping response sample data with similar trajectories in the printing cycle from the historical database, building a clamping response prediction model, and predicting the predicted response time of the pneumatic clamp in the current printing cycle;

[0061] S300, calculating a pre-trigger timing for outputting a pre-trigger instruction for the pneumatic clamp in advance based on the predicted response time and trajectory data, and then calculating a predicted threshold value of the distance between the screen printing platform and the target position;

[0062] S400, when the distance the servo axis moves to the target position is less than a prediction threshold, outputting a pre-trigger instruction for the pneumatic clamp, collecting state data during the pneumatic clamp response process, and calculating the actual response time of the pneumatic clamp based on the state data;

[0063] S500 , comparing the actual response time of the pneumatic clamp with the predicted response time to obtain a response difference value, and automatically adjusting the pre-triggering timing of the pneumatic clamp for the next printing cycle according to the response difference value.

[0064] In an embodiment of the present invention, a multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine is employed to achieve high-precision automatic coordination of the movements of the screen printing platform and the pneumatic fixture. By acquiring the positioning parameters of the target workpiece, this embodiment accurately controls the movement path of the screen printing platform, ensuring that each printing process is automatically adjusted based on the actual position of the target workpiece. The system collects and records the motion trajectory of the screen printing platform in real time, and uses this trajectory data as the core basis for all subsequent calculations and analyses, providing a solid data foundation for closed-loop optimization of the overall control strategy.

[0065] Through in-depth analysis of motion trajectory data, the system can filter out printing cycle samples from the historical database that are similar to the current working condition trajectory. These samples contain the response data of the clamping action during the actual printing process. Based on this data, the system establishes a clamping response prediction model. This model dynamically outputs the predicted response time of the pneumatic clamp based on the current actual motion state, providing a scientific estimate for subsequent synchronous actions.

[0066] During the process, the system calculates the specific timing for the pneumatic clamp to actuate based on the current predicted response time and the collected trajectory data. This information is then used to further determine the predicted threshold for the screen printing platform's distance from the target position. When the distance between the screen printing platform's current position and the target position is detected to be less than this predicted threshold, the control system proactively issues a pre-trigger command for the pneumatic clamp to ensure that the clamping action accurately matches the platform's position, minimizing synchronization errors.

[0067] During the pneumatic fixture's execution, the system synchronously collects data on the fixture's actual response status. By analyzing this data, the fixture's actual response time can be determined and compared in detail with the predicted response time. The response difference value generated by this comparison reflects the real-time synchronization accuracy of the current control system. Based on this difference value, the system automatically corrects and optimizes the pneumatic fixture's pre-trigger timing for the next printing cycle, achieving closed-loop adaptive compensation and dynamic optimization. This enables high-precision automated control of the entire screen printing process, effectively suppressing micro-displacements and quality fluctuations in the workpiece caused by asynchronous multi-axis servo and pneumatic motion.

[0068] Among them, the positioning parameters of the target workpiece are obtained, and the silk screen printing platform is driven to move by the positioning parameters, and the motion trajectory of the silk screen printing platform is collected and recorded in real time to obtain trajectory data, which specifically includes:

[0069] First, the system uses an integrated visual positioning unit, photoelectric sensors, laser ranging, or other industrial measurement methods to accurately locate the target workpiece to be printed. By identifying workpiece boundaries, reference holes, and markers, it collects numerical parameters reflecting the workpiece's spatial position and posture. These positioning parameters typically include key information such as the workpiece's X and Y coordinates and rotation angle. For example, in a PCB screen printing scenario, image recognition algorithms can automatically identify board edges and positioning holes, using their coordinates and angles as input for the current printing cycle.

[0070] After obtaining the positioning parameters, the control system inputs them into the motion control unit, which drives the screen printing platform via a multi-axis servo motor, moving the platform along a predetermined path to the set position of the target workpiece. During the platform movement process, the control system will synchronously collect the actual position information of the platform at each time point at a high frequency and record it as motion trajectory data. The trajectory data includes the spatial coordinates of the platform at each time point (such as X, Y, and Z axis positions), movement speed, acceleration, etc. The acquisition frequency can be set according to actual needs, for example, once every 10 milliseconds. All trajectory data is saved in real time in the storage module, providing detailed data support for subsequent motion analysis and dynamic control algorithms. Through this high-resolution, full-process trajectory recording method, it can be ensured that the subsequent control steps can accurately reconstruct the actual motion state of the platform at any time, effectively supporting the execution of precision printing tasks.

[0071] In a preferred embodiment of the present invention, based on the trajectory data, clamping response sample data with similar trajectories in the printing cycle are screened from the historical database, and a clamping response prediction model is constructed to predict the predicted response time of the pneumatic clamp in the current printing cycle, including:

[0072] Based on the trajectory data, the clamping response sample data with similar trajectories in the printing cycle are screened from the historical database to construct a reference data set;

[0073] Cluster analysis was performed on the clamping response time recorded in the reference dataset to extract the correlation features between the screen printing platform speed, acceleration and the pneumatic clamp response time, and generate the response time feature vector.

[0074] A clamping response prediction model is constructed based on the response time feature vector and trajectory data;

[0075] The predicted response time of the pneumatic clamp under the current working conditions is output through the clamping response prediction model.

[0076] In this embodiment of the present invention, an efficient and adaptive clamping response prediction model is constructed through in-depth analysis of motion trajectory data and clamping response samples from a historical database. In specific implementation, the system automatically selects clamping response samples from the historical database with motion trajectories similar to those of the current print cycle based on the trajectory data collected in real time during the current print cycle. This resulting reference dataset fully reflects the actual dynamic characteristics of the current operating conditions.

[0077] Furthermore, the system performs cluster analysis on the clamping response times recorded in the reference dataset, using a clustering algorithm to identify differences in response characteristics under different motion states. For example, samples with faster speeds and greater acceleration can be clustered into one category, while samples with more stable speeds and less acceleration changes can be clustered into another, thereby ensuring the targetedness and generalization capabilities of subsequent model training. After cluster analysis, the system calculates the correlation characteristics between the speed and acceleration of the screen printing platform and the response time of the pneumatic clamp under each category, forming category feature data. This data is further vectorized to obtain the response time feature vector used for model training.

[0078] Utilizing the aforementioned response time eigenvectors and current actual trajectory data, the system automatically constructs a clamping response prediction model. This model, based on a deep fit of large-sample historical data and real-time data, outputs the predicted response time of the pneumatic clamp under the specific dynamic conditions of each printing cycle in real time, providing a crucial parameter basis for the precise synchronous control of subsequent pneumatic and electrical movements. Through this adaptive modeling approach based on big data and clustering, the present invention significantly improves the robustness and practicality of the prediction model, making it particularly suitable for complex and changing real-world production environments, ensuring high efficiency and precision in printing operations.

[0079] Among them, based on the trajectory data, the clamping response sample data with similar trajectories in the printing cycle are screened from the historical database to construct a reference data set, which specifically includes:

[0080] Before each printing cycle begins, the system compares and analyzes the current platform's real-time trajectory data with the trajectory data stored in the historical database. To ensure that the selected trajectory samples are highly similar to the current working conditions, the system can use algorithms such as Euclidean distance, dynamic time warping (DTW), and correlation coefficients to quantify the similarity between the current trajectory and historical trajectories in terms of spatial path and motion timing. For example, historical trajectory samples with a small distance metric or a high correlation coefficient are selected as candidates.

[0081] For each candidate trajectory sample, the system will further compare the motion features at key nodes (such as the starting point, acceleration and deceleration sections, and platform to site), and filter out data that is significantly inconsistent with the current motion pattern. Finally, those historical clamping response samples with similar spatial paths and consistent temporal change trends are organized into a reference data set. In addition to trajectory information, each sample also contains the corresponding clamping action response time, pneumatic action parameters, etc. Taking the actual production scenario as an example, if the current printing task involves a large-size PCB, the system will give priority to filtering out historical data that has previously processed similar-sized panels, with similar motion paths and good clamping effects, to form a targeted, high-quality reference set. Through this method, the matching degree and prediction accuracy of subsequent prediction models to actual working conditions can be significantly improved, providing a solid data foundation for intelligent control.

[0082] In a preferred embodiment of the present invention, based on the predicted response time and trajectory data, the pre-trigger timing for outputting the pneumatic clamp pre-trigger instruction in advance is calculated, and then the predicted threshold value of the screen printing platform from the target position is calculated, including:

[0083] Based on the trajectory data, the speed and acceleration parameters of the screen printing platform are calculated, and the remaining running time required for the screen printing platform to move from the current position to the target position is inferred.

[0084] Calculate the pre-triggering timing for issuing the pneumatic pre-triggering command in advance based on the remaining running time and predicted response time;

[0085] According to the pre-trigger timing, combined with the current speed and acceleration of the screen printing platform, the position offset corresponding to the time point is inferred as the prediction threshold of the screen printing platform from the target position.

[0086] In an embodiment of the present invention, a set of automated methods combining real-time trajectory analysis and kinematic calculation is proposed for the calculation and synchronous control of the pre-trigger timing of the pneumatic clamp. The method is based on the position information of the platform at each sampling time point, and realizes dynamic monitoring of the motion state by continuously collecting and processing the trajectory data of the silk-screen platform. The system first analyzes the position changes and corresponding time intervals between adjacent sampling time points, calculates the speed of the silk-screen platform at each time point, and then calculates the acceleration parameters based on the speed change trend and time interval. Through the above processing, the key dynamic characteristics of the platform movement process can be obtained in real time, which provides a basis for the accurate calculation of the remaining running time.

[0087] Furthermore, based on the current speed, acceleration, and position information, the system uses kinematic principles to dynamically determine the remaining time for the platform to move from its current position to its target position. For situations where both speed and acceleration may change in actual working conditions, this method can effectively adapt to fluctuations in the motion state, ensuring that the calculation results are more consistent with the actual operating trajectory of the platform. When the platform acceleration is detected to be zero, a simplified method of dividing the distance by the speed is used to ensure computational efficiency while reducing system resource consumption. In the presence of non-zero acceleration, the system calculates the remaining running time based on the dynamic model, making the prediction results more adaptable and accurate.

[0088] By calculating the dynamic remaining operating time described above, the system combines the predicted response time to scientifically determine how long in advance the platform needs to issue a pneumatic pre-trigger instruction. Combining the current speed and acceleration, the spatial position offset corresponding to this advance timing is further inferred, and ultimately this offset is used as the predicted threshold for the platform's distance from the target position. When the platform moves within this threshold range, the control system automatically issues a pre-trigger instruction, achieving a high degree of coordination between the pneumatic fixture action and the platform movement. This embodiment can significantly improve the safety, stability, and printing quality consistency of the equipment's automated operation, meeting the needs of a high-precision manufacturing environment.

[0089] In a preferred embodiment of the present invention, the pre-triggering timing of the pneumatic clamp of the next printing cycle is automatically adjusted according to the difference value, including:

[0090] Based on the preset response tolerance threshold, calculate whether the response difference value is less than it;

[0091] When the response difference value is less than the response tolerance threshold, a proportional factor is used to perform a linear weighted correction on the response difference value to obtain a first compensation value;

[0092] When the response difference value is greater than the response tolerance threshold, a weighted average error is calculated based on the response difference values ​​recorded for a preset number of rounds, and the compensation amplitude is adjusted based on the weighted average error to obtain a second compensation value;

[0093] The first compensation value or the second compensation value is time-weightedly superimposed on the pre-trigger timing in the previous printing cycle to form a corrected pre-trigger timing, wherein the first compensation value or the second compensation value acts directly on the original trigger time as an offset.

[0094] To improve the synchronization accuracy between the pneumatic gripper's movement and the platform's motion, this embodiment of the present invention employs an adaptive compensation mechanism based on response difference values. This mechanism continuously compares the pneumatic gripper's actual response time with its predicted response time during each printing cycle, acquiring real-time data on synchronization errors. The system also sets a preset response tolerance threshold to automatically determine whether the synchronization error for the current cycle is within an acceptable range.

[0095] When the actual error is less than the tolerance threshold, a linear weighted correction is applied to the error using a proportional factor to generate a first compensation value, which directly optimizes the pre-trigger timing for the next cycle. When the actual error exceeds the tolerance threshold, the system calls up historical error information recorded over a certain number of cycles, assigns weights, and calculates the weighted average error to quantify the overall error trend. The compensation amplitude is then dynamically adjusted based on this trend to obtain a second compensation value.

[0096] Regardless of whether it is the first or second compensation value, the system performs a weighted superposition on it with the pre-trigger timing of the previous printing cycle, and finally outputs the corrected pre-trigger timing. This compensation value acts as a direct offset for timing adjustment on the control parameters, achieving dynamic closed-loop tuning. Through this compensation strategy, the system can not only quickly respond to occasional fluctuations in single errors, but also perform trend corrections for long-term error accumulation, thereby effectively eliminating synchronization offsets caused by external factors such as mechanical wear and unstable pneumatic response. The application of this embodiment greatly improves the equipment's adaptability and product consistency under long-term continuous operation conditions.

[0097] When the response difference value is less than the response tolerance threshold, a proportional factor is used to perform a linear weighted correction on the response difference value to obtain a first compensation value, specifically including:

[0098] During actual system operation, the control system continuously monitors the actual response time of the pneumatic gripper during each printing cycle and compares it with the response time predicted by the model, generating a response difference value. If the absolute value of this response difference value falls within a pre-defined acceptable range (i.e., the response tolerance threshold), the current control accuracy is determined to meet process requirements. In this case, to further fine-tune and optimize synchronization, the system introduces a proportionality factor to perform a linear weighted correction on the response difference value.

[0099] Specifically, the system sets an empirical scaling factor (e.g., 0.8) and multiplies the current response difference by this scaling factor to obtain the first compensation value. For example, if the actual response time of this round lags behind the predicted response time by 0.02 seconds, and the scaling factor is set to 0.8, the first compensation value is 0.016 seconds (0.02 seconds multiplied by 0.8). This compensation value will be directly used to correct the pre-trigger timing of the pneumatic clamp in the next printing cycle, achieving the gradual elimination of minor errors and dynamic convergence. The scaling factor can be flexibly set based on device characteristics and actual error changes, ensuring the response speed of system adjustments while avoiding new system oscillations caused by overcompensation. This linearly weighted compensation mechanism significantly improves the system's dynamic adaptability while ensuring the stability of the printing process.

[0100] In a preferred embodiment of the present invention, cluster analysis is performed on the clamping response times recorded in the reference data set to extract the correlation features between the screen printing platform speed, acceleration and the pneumatic clamp response time, and generate a response time feature vector, including:

[0101] Extract the screen printing platform speed and acceleration parameters corresponding to each clamping response time in the reference data set to form a parameter set;

[0102] Cluster the parameter set, classify samples with similar response characteristics into the same category, and obtain the sample set under each category;

[0103] For each category of sample set, the mean or median of the speed, acceleration, and response time of the category is calculated to form category feature data;

[0104] All category feature data are vectorized to form a response time feature vector containing speed, acceleration and response time statistics.

[0105] In this embodiment of the present invention, a cluster analysis method is used to automatically extract the underlying correlations between the speed and acceleration of the screen printing platform and the response time of the pneumatic clamps, addressing the complex distribution of clamping response times in a reference dataset. Specifically, the system first extracts the velocity and acceleration parameters corresponding to each clamping response sample from the reference dataset, creating a parameter set. Subsequently, a clustering algorithm is used to group samples with similar response characteristics into the same category, effectively classifying the clamping response patterns under different operating conditions.

[0106] For each category, the system calculates the mean or median of platform velocity, acceleration, and response time to generate representative category feature data. This process significantly improves the generalizability and representativeness of the sample data and helps mitigate the influence of outliers. The system then vectorizes all category feature data, ultimately generating a response time feature vector that fully reflects the response patterns of different operating conditions. This provides structured, high-information input for the subsequent construction of the clamping response prediction model.

[0107] The application of this embodiment not only improves the accuracy and robustness of subsequent clamping response prediction model training, but also provides greater adaptability to changing working conditions and complex printing scenarios. This clustering feature extraction strategy effectively improves the efficiency of the equipment's use of historical big data and significantly improves the model's generalization ability for specialized processes and batch printing tasks, providing solid data support for high-precision, intelligent printing control.

[0108] The parameter set is clustered, and samples with similar response characteristics are classified into the same category to obtain a sample set under each category, specifically including:

[0109] First, the system stores the velocity, acceleration, and response time parameters corresponding to each clamping response sample extracted from the reference dataset in a unified structure to form a parameter set. Subsequently, the system uses a clustering analysis algorithm to classify these parameters. Common clustering methods include K-means clustering, hierarchical clustering, or density-based clustering. This embodiment prefers the K-means clustering method, and the specific steps are as follows:

[0110] The system determines the required number of clusters (for example, three categories: high-speed clamping, medium-speed clamping, and low-speed clamping) based on actual production experience or preliminary experiments. It uses speed, acceleration, and response time from the entire parameter set as clustering features. The system calculates multidimensional spatial distances for all sample points based on these features, automatically grouping samples with similar response characteristics into the same category.

[0111] For example, suppose a parameter set contains 30 sample data points. After cluster analysis, these data points are divided into three groups: the first group contains samples with high speed and acceleration, the second group contains samples with moderate speed and acceleration, and the third group contains samples with low speed and low acceleration. After clustering, the samples within each group have similar dynamic characteristics and response patterns, facilitating subsequent category feature analysis and model fitting. Clustering can effectively distinguish the clamping response characteristics under different processes, batches, and parameters, providing a scientific basis for targeted optimization of model structure and parameters.

[0112] For each category of samples, the mean or median of the speed, acceleration, and response time of the category is calculated to form category feature data, including:

[0113] After clustering is complete, the system performs statistical analysis on the three parameters of speed, acceleration, and response time for all samples in each category. Statistical analysis methods primarily include calculating the arithmetic mean (mean) and the median. For categories with a large number of samples and a relatively even data distribution, it is recommended to use the mean as the statistical result. For categories with outliers or skewed distributions, the median can be used to reduce interference from extreme values. For example, if a cluster category contains 10 samples with speeds of 100, 102, 99, etc., the system will add up all speed parameters and divide by the number of samples to obtain the mean speed for that category. The same method is used for statistics on acceleration and response time.

[0114] If the influence of noise and outliers needs to be further eliminated, the system can pre-eliminate extreme outliers and then perform mean or median statistics. Ultimately, the system generates a set of category feature data for each category, which reflects the typical dynamic response characteristics of samples under that category. In the subsequent model training and prediction stages, the system will use these category feature data as core input to improve the pertinence and accuracy of the clamping response prediction model. For example, if the mean velocity of a certain category is 105mm / s, the mean acceleration is 12mm / s², and the median response time is 0.18 seconds, then this set of data is the category feature data for that category, which can be directly used to describe and predict the response behavior under subsequent similar working conditions.

[0115] In a preferred embodiment of the present invention, a clamping response prediction model is constructed based on the response time feature vector and trajectory data, including:

[0116] The response time characteristic vector and the trajectory data collected during the current printing cycle are used as input parameters to determine the input variables;

[0117] Based on the relationship between historical input variables and their corresponding actual response time of the pneumatic clamp, the mapping relationship between the two is parameter fitted to establish a clamping response prediction model;

[0118] The prediction accuracy of the clamping response prediction model is verified by comparing it with the actual response time of the pneumatic clamp collected in the current or subsequent printing cycle. The model parameters are corrected according to the comparison results to obtain the corrected clamping response prediction model.

[0119] In this embodiment of the present invention, the construction and adaptive updating of the clamping response prediction model are further improved by incorporating response time eigenvectors and trajectory data. During implementation, the system uses the response time eigenvectors obtained through cluster analysis and trajectory data collected during the current printing cycle as input variables to build the clamping response prediction model. This approach automatically combines the changing characteristics of historical and real-time data to dynamically model and accurately predict the response behavior of pneumatic clamps under different operating conditions.

[0120] In actual applications, the system performs parameter fitting based on the relationship between historical input variables and the corresponding actual response times of the pneumatic grippers. Fitting methods can employ common engineering algorithms such as minimum mean square error and regression analysis to efficiently approximate the response patterns. To ensure the model's effectiveness in practical applications, the system also verifies the accuracy of the clamping response prediction model after the model is established. Specifically, the system compares the model's predictions with the actual response times of the pneumatic grippers collected during the current or subsequent printing cycles. If significant errors occur, the model parameters are automatically corrected and adaptively optimized.

[0121] Through this closed-loop model adaptation mechanism, the clamping response prediction model can continuously adapt to external disturbances such as equipment status, environmental changes, and material variations during the actual production process, maintaining high prediction accuracy and stability. This embodiment not only ensures the model's engineering practicality and real-time performance, but also significantly enhances the automation system's self-learning and intelligent adjustment capabilities in complex manufacturing environments, providing a strong guarantee for efficient and stable linkage control of fully automatic screen printing machines.

[0122] Among them, based on the relationship between historical input variables and their corresponding actual response time of the pneumatic clamp, parameter fitting is performed on the mapping relationship between the two to establish a clamping response prediction model, which specifically includes:

[0123] During implementation, the system first uses the response time feature vectors obtained through clustering and statistical processing, along with motion trajectory data collected from each historical printing cycle, as input variables. These input variables are then paired with the actual pneumatic clamp response times recorded in the historical database to form training data pairs. To construct a clamping response prediction model, the system uses parameter fitting to model the relationship between the input variables and the actual response times. Common parameter fitting methods include multivariate linear regression, nonlinear least squares, or machine learning regression algorithms (such as decision tree regression and random forest regression).

[0124] During specific operations, the system will establish a data set based on the input variables (such as platform speed, acceleration, category characteristics, etc.) and the target output (i.e., the corresponding actual response time), and use a fitting algorithm to determine the influence weight of each input variable on the response time and the model structure. For example, when using linear regression, the system adjusts the model parameters to minimize the error between the predicted response time and the actual response time of all samples. If a machine learning regression method is used, the system repeatedly trains and corrects the model parameters so that the model can automatically capture the nonlinear laws of the response time under different combinations of input variables. Through the above method, the system ultimately obtains a clamping response prediction model that can input real-time working condition variables and output the corresponding pneumatic clamp response time, providing a reliable prediction basis for subsequent action synchronization and compensation.

[0125] The prediction accuracy of the clamping response prediction model is verified by comparing it with the actual response time of the pneumatic clamp collected in the current or subsequent printing cycle. The model parameters are corrected according to the comparison results to obtain the corrected clamping response prediction model. Specifically, the following steps are performed:

[0126] After the initial model establishment, the system enters the prediction accuracy verification and adaptive correction phase. Specifically, during actual production, the system utilizes the trained clamping response prediction model to predict the pneumatic clamp response time for the current print cycle based on the input variables (such as the most recently collected response time feature vector and trajectory data). The system then measures and records the actual pneumatic clamp response time. The system then compares the model predictions with the actual values ​​to determine the model's prediction error.

[0127] To ensure high model accuracy over the long term, the system automatically optimizes model parameters using an adaptive parameter correction strategy based on error size and distribution. For example, this can fine-tune the weight parameters of a linear regression model or incrementally adjust the parameters of branches and leaf nodes in a machine learning regression model. This correction can employ minimum error feedback adjustment, whereby when the error exceeds a set threshold, the system automatically adjusts the model parameters to reduce the gap between the predicted and actual values.

[0128] For example, if the model continuously predicts slow response times within a production batch, the system can automatically increase the weight of the speed or acceleration parameters in the input variables to improve the model's sensitivity to response times under fast conditions. By continuously repeating the "prediction-verification-correction" process, the system achieves continuous self-learning and intelligent evolution of the clamping response prediction model, enabling it to consistently adapt to changes in equipment status, fluctuations in material batches, or external environmental interference, ensuring the long-term stability of printing control accuracy and synchronization performance.

[0129] In a preferred embodiment of the present invention, the velocity and acceleration parameters of the screen printing platform are calculated based on the trajectory data, and the remaining running time required for the screen printing platform to move from the current position to the target position is inferred, including:

[0130] According to the trajectory data, the speed of the screen printing platform at the current time point is obtained by calculating the position change and time interval between adjacent sampling time points;

[0131] Calculate the velocity change and time interval of adjacent sampling time points to obtain the acceleration of the screen printing platform at the current time point;

[0132] Based on the position information, speed and acceleration parameters of the screen printing platform at each time point, the distance between the current position of the screen printing platform and the target position is determined. Based on the principles of kinematics, the remaining running time required for the screen printing platform to move to the target position at the current speed and acceleration is determined.

[0133] When it is detected that the acceleration of the screen printing platform is zero, the distance between the current position and the target position is divided by the current speed to obtain the remaining running time required to move to the target position at a constant speed.

[0134] In an embodiment of the present invention, the velocity and acceleration parameters of the screen printing platform are dynamically calculated based on trajectory data, and the remaining travel time required for the platform to move from its current position to its target position is inferred from this data, achieving high-precision timing prediction in a multi-axis servo control system. During implementation, the system collects platform position and time information at each sampling point, calculates the position change and time interval between adjacent sampling points, and obtains real-time platform velocity data at each point in time. Furthermore, the system processes adjacent velocity changes and corresponding time intervals to obtain the platform's acceleration parameters at each point in time.

[0135] This method allows the system to dynamically characterize the entire motion of the screen printing platform at all points in time, ensuring that every control decision is based on the latest and most realistic physical state data. Using known velocity, acceleration, and current position, the system calculates the remaining time required for the platform to reach its target position based on kinematic principles. For zero acceleration, the system quickly calculates the remaining time using a simplified method: dividing distance by velocity, thereby improving data processing efficiency. For dynamic conditions with non-zero acceleration, the system combines velocity and acceleration trends to accurately estimate the remaining time using a derived formula or iterative algorithm.

[0136] This remaining run time prediction based on real-time dynamic data provides solid time parameter support for subsequent pre-trigger control of the pneumatic fixture. Practical application results demonstrate that this embodiment effectively improves the system's dynamic response and control accuracy, significantly reducing the problem of asynchrony between the fixture and the platform caused by inaccurate motion state estimation. This ensures efficient operation of the automatic screen printing equipment while achieving stable process quality output.

[0137] The process involves determining the distance between the current position of the screen printing platform and the target position based on the position information, speed, and acceleration parameters of the screen printing platform at each time point. Furthermore, based on the principles of kinematics, the remaining time required for the screen printing platform to move to the target position at the current speed and acceleration is determined. Specifically, the process involves:

[0138] First, the control system samples the screen printing platform's motion at a high frequency, synchronously recording its actual position (such as X, Y, and Z coordinates), current velocity, and acceleration parameters at each point in time. The system then compares the platform's current coordinates with the set target position and calculates the linear distance between them. For example, if the target position on the X axis is 500mm and the current position is 350mm, the distance is 150mm.

[0139] After determining the distance, the system combines the current speed and acceleration and uses the principles of kinematics to calculate the remaining running time required for the platform to reach the target position. For the case where the acceleration is zero, the system directly calculates by dividing the distance by the speed; when the acceleration is not zero, the system follows the law of uniform acceleration linear motion, based on the known speed, acceleration and distance, and uses the method of solving a quadratic equation to determine how long it will take for the platform to reach the target position. Since actual control systems tend to implement such operations directly in the program, the above process can be converted into a series of conditional judgments and numerical iterations through programming during specific operations, such as using the Newton iteration method or the step-by-step interpolation method to gradually approach the accurate remaining time. Ultimately, the system uses the calculated remaining running time as an important parameter input for the timing decision of the pneumatic clamp action, providing accurate time base support for subsequent collaborative control strategies.

[0140] For example,

[0141] Assuming the platform is currently 100 mm from the target position, its velocity is 200 mm / s, and its acceleration is -50 mm / s², the system can loop through the program to gradually calculate how long it will take for the platform to decelerate to zero and reach the target point, thereby estimating the remaining run time.

[0142] In a preferred embodiment of the present invention, calculating the pre-triggering timing for issuing the pneumatic pre-triggering instruction in advance based on the remaining operating time and the predicted response time includes:

[0143] By comparing the remaining running time with the predicted response time, it is determined that when the remaining running time of the screen printing platform is earlier than the predicted response time, a pre-trigger instruction of the pneumatic clamp is issued in advance;

[0144] Subtract the predicted response time from the remaining running time to calculate the specific value of the pre-trigger instruction and obtain the pre-trigger timing at which the screen printing platform should issue the pre-trigger instruction in advance;

[0145] When the pre-trigger timing is less than the preset minimum safety advance value, the minimum safety advance value is used as the pre-trigger timing for issuing the pre-trigger instruction, and the pre-trigger timing is output.

[0146] In this embodiment of the present invention, a method for calculating the pneumatic clamp pre-trigger timing based on a combination of remaining run time and predicted response time is employed, achieving highly coordinated coordination across multiple steps in the printing process. After obtaining the remaining run time of the screen printing platform and the predicted response time of the pneumatic clamp, the system compares these two time parameters to determine how far in advance the platform should issue the pneumatic clamp pre-trigger command. By subtracting the predicted response time from the remaining run time, the system calculates the specific time at which the platform should issue the pre-trigger command, thereby enabling intelligent scheduling of the pneumatic clamp's action sequence.

[0147] This process also incorporates a preset minimum safety lead time. This ensures that, in actual operating conditions, even if the calculated pre-trigger timing is less than the safety limit, the system prioritizes safety and automatically uses the minimum safety lead time, thus preventing clamping failure or product damage caused by premature or late triggering. The system ultimately outputs the determined pre-trigger timing as an action control parameter for subsequent spatial determination and action delivery.

[0148] This implementation not only optimizes the matching between the pneumatic fixture's motion response and the dynamic operation of the screen printing platform, minimizing the adverse effects of synchronization errors, but also significantly improves the safety and robustness of the printing automation system under complex process flows. In practical applications, this method can help users flexibly adapt to production scenarios with varying speeds and response characteristics, enabling fully automatic screen printing equipment to maintain excellent synchronization control performance in high-intensity production environments.

[0149] The method for setting the minimum safety advance value specifically includes:

[0150] In the pneumatic clamp motion control process, to avoid the clamping action not being completed in time due to the pre-trigger timing being set too late, or synchronization failure due to error fluctuations, the system needs to set a "minimum safety advance value" as the minimum advance time guarantee for the pre-trigger action. The setting of the minimum safety advance value should be comprehensively evaluated based on the physical response characteristics of the pneumatic clamp, equipment inertia, process safety requirements, and historical production experience. Common setting methods include:

[0151] First, the system uses the equipment's technical specifications. The system can query the pneumatic clamp manufacturer's datasheet to determine the maximum response time required for the clamp to complete the clamping action at standard pressure. For example, if a certain clamp has a maximum response time of 0.12 seconds at standard operating pressure, a safety margin of 10-20% can be added to this time, resulting in a minimum safety lead time of 0.14 seconds.

[0152] The second approach is to integrate historical statistical data. The system can collect data on the actual response times of pneumatic fixtures during mass production, sort all sampled response time data, select the 99th percentile or maximum value, and add a margin to set it as the safety lead threshold. For example, if 99% of the response times across thousands of actual response acquisitions are less than 0.15 seconds, the minimum safety lead value can be set to 0.16 seconds.

[0153] Third, based on process safety redundancy requirements, if the actual production process has high requirements for clamping action synchronization, a more conservative redundancy time can be used to ensure that the action can be completed on time even under extreme working conditions, avoiding risks to printing quality or equipment safety.

[0154] In practical applications, it's recommended that the system be initially configured using a combination of theory and experience. Subsequently, the minimum safety lead value can be dynamically adjusted through online monitoring and actual performance feedback. For example, if a delay in the clamping action is detected, the minimum safety lead value can be automatically adjusted upwards. This approach ensures that sufficient buffer time is always provided when the pre-trigger command is issued, safeguarding equipment operation safety and consistent product quality.

[0155] In a preferred embodiment of the present invention, when the response difference value is greater than the response tolerance threshold, a weighted average error is calculated based on the response difference values ​​recorded for a preset number of rounds, and the compensation amplitude is adjusted based on the weighted average error to obtain a second compensation value, including:

[0156] The response difference values ​​generated in the current printing cycle are combined with the response difference values ​​recorded in a preset number of rounds to form an error processing data set;

[0157] Based on the weight distribution principle set in descending order according to the distance from the current printing cycle, a weight coefficient is assigned to each response difference value in the error processing data set;

[0158] Perform weighted summation on all response difference values ​​that have been assigned weight coefficients, and divide the weighted summation value by the sum of the weight coefficients to obtain the weighted average error used for compensation adjustment;

[0159] According to the absolute value and change trend of the weighted average error, the compensation amplitude used to correct the pre-trigger timing of the next cycle is adjusted so that the compensation amplitude automatically adapts to the cumulative error trend to generate a second compensation value.

[0160] In this embodiment of the present invention, to address the fluctuations in pneumatic gripper response errors that can occur during actual production, the system incorporates a weighted average error and adaptive compensation amplitude adjustment mechanism based on historical multi-round response difference values. In practice, the system collects the response difference values ​​generated by the current printing cycle in real time and combines them with the response difference values ​​recorded from previous printing cycles to form a complete error processing dataset. This data set truly reflects the synchronization error dynamics of the equipment during recent continuous operation, facilitating comprehensive analysis of error trends.

[0161] To ensure realistic error compensation, the system employs a weighting principle that descends sequentially based on proximity to the current printing cycle. Specifically, response differences closer to the current cycle are assigned higher weights, while errors from earlier cycles are assigned lower weights. This allows the compensation strategy to more sensitively respond to recent synchronization shifts. The system then takes the weighted sum of all weighted response differences and divides the result by the sum of the weighting factors to calculate the weighted average error. By continuously analyzing the absolute value of the weighted average error and its cycle-by-cycle trend, it accurately captures both the long-term cumulative effects of error changes and occasional fluctuations.

[0162] Based on this, the system automatically adjusts the compensation amplitude for the next cycle's pre-trigger timing correction based on the magnitude and changing trend of the weighted average error, generating a second compensation value. This not only enables immediate correction of single errors but also adaptive amplitude adjustments based on long-term error trends, significantly improving the printing equipment's long-term adaptability to complex influencing factors such as environmental fluctuations, mechanical wear, and pneumatic system aging. Practice has proven that this mechanism effectively prevents the accumulation and expansion of synchronization errors, improving equipment stability and process consistency during high-rate, long-running continuous production.

[0163] The compensation amplitude for correcting the pre-trigger timing of the next cycle is adjusted according to the absolute value and change trend of the weighted average error, so that the compensation amplitude automatically adapts to the cumulative error trend to generate a second compensation value, specifically including:

[0164] First, the system calculates a weighted average error by weighting the difference in responses recorded during previous printing cycles against the difference in responses during the current cycle. To make the compensation strategy more intelligent and sensitive, the system continuously monitors the absolute value of this weighted average error and its trend over time. For example, if the weighted average error is consistently positive and gradually increases, it indicates that the pneumatic gripper's pre-triggering is generally late, requiring increased compensation to bring it into action sooner. If the error gradually decreases or even reverses, the compensation intensity can be reduced accordingly to prevent over-compensation.

[0165] The specific adjustment steps are:

[0166] The system first sets a baseline amplitude and adjustment step size. When the absolute value of the weighted average error is less than a predetermined safety threshold, the compensation amplitude remains unchanged or is fine-tuned. When the absolute value exceeds the threshold, the system adaptively increases or decreases the compensation amplitude based on the rate of error change (e.g., a continuously increasing or decreasing trend). For example, if the weighted average error over five rounds is 0.01, 0.02, 0.03, 0.04, and 0.06 seconds, respectively, the system identifies a continuously increasing trend and increases the compensation amplitude by 1.2 times the original amplitude for the next round. Conversely, if the error over the last few rounds has stabilized or reversed, the compensation factor can be automatically reduced or gradually converged to a baseline value. The system can dynamically analyze error trends using algorithms such as sliding windows and moving averages, allowing compensation to adapt to process needs in a timely and automatic manner based on the cumulative error trend. This strategy ensures a sensitive and stable compensation response, avoiding system jitter or non-convergence caused by sudden or abnormal changes.

[0167] In a preferred embodiment of the present invention, the first compensation value or the second compensation value is time-weightedly superimposed on the pre-trigger timing in the previous printing cycle to form a corrected pre-trigger timing, including:

[0168] The first compensation value or the second compensation value in the current printing cycle is used as input data together with the pre-trigger timing recorded in the previous printing cycle;

[0169] Determining a weighting coefficient for the current printing cycle according to the first compensation value or the second compensation value in the current printing cycle, a historical error change trend, and a preset correction priority;

[0170] The first compensation value or the second compensation value is multiplied by a weighting coefficient to obtain a weighted compensation amount, and the weighted compensation amount is added to the pre-trigger timing in the previous printing cycle to obtain a corrected pre-trigger timing.

[0171] In this embodiment of the present invention, a compensation correction strategy based on time-weighted superposition is employed to dynamically optimize the pre-trigger timing of the pneumatic clamp. During implementation, the system inputs the first or second compensation value obtained during the current print cycle, along with the pre-trigger timing recorded during the previous print cycle, into the compensation calculation module. The system combines the current compensation value, historical error trends, and pre-set correction priorities to determine a weighting coefficient for the current cycle. This coefficient automatically adjusts based on the actual error response sensitivity, ensuring that the correction strategy is neither overly drastic nor sluggish.

[0172] Next, the system multiplies the first or second compensation value by the weighting coefficient to obtain a weighted compensation amount. This weighted compensation amount is then used as an offset and applied directly to the pre-trigger timing in the previous printing cycle. Through time-weighted superposition, a corrected pre-trigger timing is obtained. The system uses this correction result to control the timing of the pneumatic gripper's movement in the next cycle, ensuring continuous and adaptive compensation.

[0173] By introducing time-weighted and multi-factor superposition, this implementation effectively balances the ongoing impact of historical errors on system tuning with the immediate response to the latest errors, making the entire compensation process more intelligent and flexible. Practical applications have demonstrated that this compensation strategy can rapidly suppress response timing drift caused by various disturbances, such as production batches, ambient temperature, and pneumatic pressure, while continuously optimizing fixture synchronization and significantly improving the long-term operational stability and product consistency of screen printing automation equipment in complex and changing environments.

[0174] The steps for setting the correction priority include:

[0175] Factors that influence the prioritization of revisions:

[0176] (1) The absolute value of the actual response error of the current printing cycle;

[0177] (2) The changing trend of errors in several historical cycles (e.g., the error continues to increase, decrease, or fluctuate);

[0178] (3) Current process stage (such as first-time operation, material switching, restart after equipment maintenance, and other critical periods);

[0179] (4) The quality requirement level set by the user or system (e.g., precision electronics and high-end printing have higher priority).

[0180] Error classification and grading: The system sets grading thresholds based on the current absolute value of the error. For example, less than 0.01 seconds is set as "low", 0.01-0.03 seconds is set as "medium", and greater than 0.03 seconds is set as "high". The system automatically determines the current error grade.

[0181] Trend analysis empowerment: Perform sliding window statistics on historical errors. If the error continues to increase for several consecutive rounds (such as 3 to 5 rounds), the priority will be automatically increased; if the error tends to be stable or continues to decrease, the priority will be lowered.

[0182] Adjustments are made based on the process stage and quality requirements: When switching to a new batch, a new process, or performing high-precision tasks, the system automatically sets the correction priority to high; under standard working conditions or when the error fluctuation is slight, it is set to medium or low.

[0183] Priority quantification and output: Convert the corrected priority into a specific value, such as high priority = 1.0, medium = 0.7, and low = 0.4. The system outputs the final priority quantization value to the compensation module for weighted coefficient calculation.

[0184] The weighting coefficient of the current printing cycle is determined according to the first compensation value or the second compensation value in the current printing cycle, the historical error change trend, and the preset correction priority, specifically including:

[0185] Collect input variables: The system obtains in real time the first compensation value or the second compensation value of the current printing cycle, the trend analysis results of the historical N rounds of errors (such as the latest 5 rounds), and the correction priority value just quantified and output.

[0186] Trend correction factor calculation: Perform a linear fit or moving average on historical errors. If the trend is continuously increasing, set the trend correction factor to >1 (e.g., 1.2); if the trend is stable, set it to 1.0; if the trend is decreasing, it can be appropriately lower than 1 (e.g., 0.8). This factor is used to amplify or converge the weight of the immediate compensation value.

[0187] Initial setting of instant compensation weight: Set the instant compensation weight base (such as 0.5~0.8) to preliminarily assign the intensity of the compensation amount participating in the weighting.

[0188] Calculating the weighting coefficient: The system multiplies the correction priority, trend correction factor, and immediate compensation weight to obtain the final weighting coefficient. For example, the weighting coefficient = correction priority value × trend correction factor × immediate compensation weight.

[0189] Weighting coefficient calibration and boundary constraints: To prevent abnormal fluctuations, the system sets the maximum and minimum values ​​of the weighting coefficient, such as the maximum not exceeding 1.0 and the minimum not less than 0.1, to ensure that the system adjustment is both sensitive and robust.

[0190] Output weighting coefficient and participate in compensation calculation: The calculated weighting coefficient is directly used for the weighting operation of the first compensation value or the second compensation value, and participates in the pre-trigger timing correction of the current cycle.

[0191] For example:

[0192] Assume that the error of the current cycle is 0.035 seconds, the historical trend is that the error has gradually increased in the past five cycles, and the process is a high-precision task:

[0193] Correction priority (high): 1.0, trend correction factor (continuously increasing): 1.2, immediate compensation weight (experience): 0.8, then the weighting coefficient = 1.0×1.2×0.8=0.96, if the second compensation value of this round is 0.03 seconds, then the weighted compensation amount is 0.0288 seconds.

[0194] The first compensation value or the second compensation value is multiplied by a weighting coefficient to obtain a weighted compensation amount, and the weighted compensation amount is added to the pre-trigger timing in the previous printing cycle to obtain a corrected pre-trigger timing, specifically including:

[0195] After the system obtains the weighting coefficient, it directly multiplies the first or second compensation value of the current round by the weighting coefficient to obtain the actual weighted compensation amount. For example, if the current round compensation value is 0.01 seconds and the weighting coefficient is 0.8, the weighted compensation amount is 0.008 seconds.

[0196] The system then uses the pre-trigger timing of the previous print cycle as a reference, directly adding or subtracting the calculated weighted compensation amount (depending on the compensation direction) as a time offset to determine the corrected pre-trigger timing for the current print cycle. For example, if the pre-trigger timing of the previous print cycle was 0.10 seconds before the platform reached its designated position, and the weighted compensation for this print cycle is 0.008 seconds, the corrected pre-trigger timing is 0.108 seconds. This new timing serves as the reference parameter for the next print cycle and is automatically input into the pneumatic control unit, enabling closed-loop, progressive, adaptive dynamic adjustment.

[0197] This approach ensures that each compensation adjustment is gradual and controlled, preventing significant system fluctuations caused by a single anomaly. It also balances current control accuracy with the self-optimization capabilities of historical trends. Through continuous superposition and convergence, it ultimately achieves long-term, high-precision synchronization between the platform and pneumatic gripper, improving product yield and production consistency.

[0198] An embodiment of the present invention further provides a multi-axis servo linkage and pneumatic-electrical coordinated control system for a fully automatic screen printing machine, the system comprising:

[0199] Positioning parameter acquisition module, used to obtain the positioning parameters of the target workpiece;

[0200] The platform drive and trajectory acquisition module is used to drive the screen printing platform to move according to the positioning parameters, and to collect and record the motion trajectory of the screen printing platform in real time to obtain trajectory data;

[0201] The similar sample screening and prediction model building module is used to screen the clamping response sample data with similar trajectories in the printing cycle from the historical database based on the trajectory data, and build a clamping response prediction model to predict the predicted response time of the pneumatic clamp under the current printing cycle;

[0202] The pre-trigger timing calculation module is used to calculate the pre-trigger timing for outputting the pneumatic fixture pre-trigger instruction in advance based on the predicted response time and trajectory data, and then calculate the predicted threshold value of the screen printing platform from the target position;

[0203] The pneumatic motion control module is used to output a pre-trigger instruction for the pneumatic fixture when the distance between the screen printing platform and the target position is less than the predicted threshold, collect the status data of the pneumatic fixture during the response process, and calculate the actual response time of the pneumatic fixture based on the status data;

[0204] The pre-trigger timing compensation module is used to compare the actual response time of the pneumatic clamp with the predicted response time to obtain a response difference value, and automatically adjust the pre-trigger timing of the pneumatic clamp for the next printing cycle according to the response difference value.

[0205] It should be noted that this system is a system corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effects.

[0206] An embodiment of the present invention further provides a computing device comprising: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the above-described method. All implementations in the above-described method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0207] The embodiment of the present invention further provides a computer-readable storage medium storing instructions, which, when executed on a computer, causes the computer to execute the above-described method. All implementations in the above-described method embodiment are applicable to this embodiment and can achieve the same technical effects.

[0208] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. The multi-axis servo linkage and pneumatic-electric coordinated control method of the fully automatic screen printing machine is characterized by: The method comprises: Obtain the positioning parameters of the target workpiece, and use them to drive the screen printing platform to move, collect and record the motion trajectory of the screen printing platform in real time, and obtain trajectory data; Based on the trajectory data, the clamping response sample data with similar trajectories in the printing cycle are screened from the historical database, and a clamping response prediction model is constructed to predict the predicted response time of the pneumatic clamp under the current printing cycle; Based on the predicted response time and trajectory data, the pre-trigger timing for outputting the pneumatic fixture pre-trigger command in advance is calculated, and then the predicted threshold of the screen printing platform's distance from the target position is calculated; When the distance the servo axis moves to the target position is less than the prediction threshold, a pre-trigger instruction for the pneumatic fixture is output, the status data of the pneumatic fixture during the response process is collected, and the actual response time of the pneumatic fixture is calculated based on the status data; The actual response time of the pneumatic clamp is compared with the predicted response time to obtain a response difference value, and the pre-trigger timing of the pneumatic clamp for the next printing cycle is automatically adjusted according to the response difference value.

2. The multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine according to claim 1, characterized in that: Based on the trajectory data, the clamping response sample data with similar trajectories in the printing cycle are screened from the historical database, and a clamping response prediction model is constructed to predict the predicted response time of the pneumatic clamp under the current printing cycle, including: Based on the trajectory data, the clamping response sample data with similar trajectories in the printing cycle are screened from the historical database to construct a reference data set; Cluster analysis was performed on the clamping response time recorded in the reference dataset to extract the correlation features between the screen printing platform speed, acceleration and the pneumatic clamp response time, and generate the response time feature vector. A clamping response prediction model is constructed based on the response time feature vector and trajectory data; The predicted response time of the pneumatic clamp under the current working conditions is output through the clamping response prediction model.

3. The multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine according to claim 1, characterized in that: Based on the predicted response time and trajectory data, the pre-trigger timing for outputting the pneumatic fixture pre-trigger command in advance is calculated, and then the predicted threshold of the screen printing platform's distance from the target position is calculated, including: Based on the trajectory data, the speed and acceleration parameters of the screen printing platform are calculated, and the remaining running time required for the screen printing platform to move from the current position to the target position is inferred. Calculate the pre-triggering timing for issuing the pneumatic pre-triggering command in advance based on the remaining running time and predicted response time; According to the pre-trigger timing, combined with the current speed and acceleration of the screen printing platform, the position offset corresponding to the pre-trigger timing is inferred as the prediction threshold of the screen printing platform from the target position.

4. The multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine according to claim 1, characterized in that: Automatically adjust the pre-trigger timing of the pneumatic clamp for the next printing cycle based on the response difference value, including: Calculate whether the response difference value is less than the preset response tolerance threshold; When the response difference value is less than the response tolerance threshold, a proportional factor is used to perform a linear weighted correction on the response difference value to obtain a first compensation value; When the response difference value is greater than the response tolerance threshold, a weighted average error is calculated based on the response difference values ​​recorded for a preset number of rounds, and the compensation amplitude is adjusted based on the weighted average error to obtain a second compensation value; The first compensation value or the second compensation value is time-weightedly superimposed on the pre-trigger timing in the previous printing cycle to form a corrected pre-trigger timing, wherein the first compensation value or the second compensation value acts directly on the original trigger time as an offset.

5. The multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine according to claim 2, characterized in that: Cluster analysis is performed on the clamping response times recorded in the reference dataset to extract the correlation features between the screen printing platform speed, acceleration and the pneumatic clamp response time, and generate a response time feature vector, including: Extract the screen printing platform speed and acceleration parameters corresponding to each clamping response time in the reference data set to form a parameter set; Cluster the parameter set, classify samples with similar response characteristics into the same category, and obtain the sample set under each category; For each category of sample set, the mean or median of the speed, acceleration, and response time of the category is calculated to form category feature data; All category feature data are vectorized to form a response time feature vector containing speed, acceleration and response time statistics.

6. The multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine according to claim 5, characterized in that: Based on the response time feature vector and trajectory data, a clamping response prediction model is constructed, including: The response time characteristic vector and the trajectory data collected during the current printing cycle are used as input parameters to determine the input variables; Based on the relationship between historical input variables and their corresponding actual response time of the pneumatic clamp, the mapping relationship between the two is parameter fitted to establish a clamping response prediction model; The prediction accuracy of the clamping response prediction model is verified by comparing it with the actual response time of the pneumatic clamp collected in the current or subsequent printing cycle. The model parameters are corrected according to the comparison results to obtain the corrected clamping response prediction model.

7. The multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine according to claim 3, characterized in that: Based on the trajectory data, calculate the speed and acceleration parameters of the screen printing platform, and infer the remaining running time required for the screen printing platform to move from the current position to the target position, including: According to the trajectory data, the speed of the screen printing platform at the current time point is obtained by calculating the position change and time interval between adjacent sampling time points; Calculate the velocity change and time interval of adjacent sampling time points to obtain the acceleration of the screen printing platform at the current time point; Based on the position information, speed and acceleration parameters of the screen printing platform at each time point, the distance between the current position of the screen printing platform and the target position is determined. Based on the principles of kinematics, the remaining running time required for the screen printing platform to move to the target position at the current speed and acceleration is determined. When it is detected that the acceleration of the screen printing platform is zero, the distance between the current position and the target position is divided by the current speed to obtain the remaining running time required to move to the target position at a constant speed.

8. The multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine according to claim 7, characterized in that: Based on the remaining running time and the predicted response time, the pre-trigger timing for issuing the pneumatic pre-trigger command is calculated in advance, including: By comparing the remaining running time with the predicted response time, it is determined that when the remaining running time of the screen printing platform is earlier than the predicted response time, a pre-trigger instruction of the pneumatic clamp is issued in advance; Subtract the predicted response time from the remaining running time to calculate the specific value of the pre-trigger instruction and obtain the pre-trigger timing at which the screen printing platform should issue the pre-trigger instruction in advance; When the pre-trigger timing is less than the preset minimum safety advance value, the minimum safety advance value is used as the pre-trigger timing for issuing the pre-trigger instruction, and the pre-trigger timing is output.

9. The multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine according to claim 4, characterized in that: When the response difference value is greater than the response tolerance threshold, a weighted average error is calculated based on the response difference values ​​recorded for a preset number of rounds, and the compensation amplitude is adjusted based on the weighted average error to obtain a second compensation value, including: The response difference values ​​generated in the current printing cycle are combined with the response difference values ​​recorded in a preset number of rounds to form an error processing data set; Based on the weight distribution principle set in descending order according to the distance from the current printing cycle, a weight coefficient is assigned to each response difference value in the error processing data set; Perform weighted summation on all response difference values ​​that have been assigned weight coefficients, and divide the weighted summation value by the sum of the weight coefficients to obtain the weighted average error used for compensation adjustment; According to the absolute value and change trend of the weighted average error, the compensation amplitude used to correct the pre-trigger timing of the next cycle is adjusted so that the compensation amplitude automatically adapts to the cumulative error trend to generate a second compensation value.

10. The multi-axis servo linkage and pneumatic-electrical coordinated control method for a fully automatic screen printing machine according to claim 9, characterized in that: Performing time-weighted superposition of the first compensation value or the second compensation value and the pre-trigger timing in the previous printing cycle to form a corrected pre-trigger timing, including: The first compensation value or the second compensation value in the current printing cycle is used as input data together with the pre-trigger timing recorded in the previous printing cycle; Determining a weighting coefficient for the current printing cycle according to the first compensation value or the second compensation value in the current printing cycle, a historical error change trend, and a preset correction priority; The first compensation value or the second compensation value is multiplied by a weighting coefficient to obtain a weighted compensation amount, and the weighted compensation amount is added to the pre-trigger timing in the previous printing cycle to obtain a corrected pre-trigger timing.

Images