Friction welding machine sliding table control method and device

Abstract

The invention relates to the technical field of friction welding machines, and discloses a friction welding machine sliding table control method and device.The method comprises the following steps that movement data of a friction welding machine stirring head are collected in real time, and a stirring head kinetic model is established; based on the mobile data and the dynamic model, outputting a preliminary control strategy through an optimal linear quadratic regulation control algorithm; on the basis of the movement data, future movement data of the stirring head are predicted through a prediction control algorithm, and a preliminary control strategy is optimized; a control signal is sent to a servo motor of the sliding table to control the servo motor; and the difference between the real-time movement data of the stirring head and the prediction result is used as feedback data, and control strategy parameters are adjusted. According to the method, the difference between the real-time movement data of the stirring head and the prediction result is used as feedback, a closed-loop control mechanism is formed, and control parameters are continuously monitored and optimized. The position control precision of the stirring head is improved, the problem of response lag in high-speed welding is effectively solved, the consistency of welding line tracks is ensured, and position deviation is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of friction welding machines, and in particular to a method and device for controlling a slide table of a friction welding machine. Background Art

[0002] Currently in the welding industry, especially for industrial applications that require high-precision and high-strength connections, friction welding technology has received extensive attention due to its unique advantages. However, the existing friction welding machine control systems have deficiencies in the accuracy of displacement control, resulting in an inaccurate working travel trajectory of the stirring head during the welding process, thus affecting the quality of the weld seam.

[0003] Existing friction welding machines usually adopt open-loop or simple closed-loop control systems to control the displacement of the slide table to drive the stirring head for welding work. Due to the lack of precise feedback and control mechanisms, the above-mentioned existing technical solutions are difficult to ensure the stability and accuracy of the stirring head during the welding process. Especially during high-speed welding and welding of complex structural parts, problems such as uneven weld seams and unstable welding strength are likely to occur. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and device for controlling a slide table of a friction welding machine, which solves the problems in the prior art that friction welding machines usually adopt open-loop or simple closed-loop control systems, lack precise feedback and control mechanisms, and are prone to uneven weld seams and unstable welding strength.

[0005] To achieve the above object, the present invention is realized through the following technical solutions: A method for controlling a slide table of a friction welding machine includes the following steps: Real-time collect the movement data of the stirring head of the friction welding machine, the movement data includes the movement direction and the movement speed, and establish a dynamic model of the stirring head; Based on the movement data and the dynamic model, calculate the optimal control input through the optimal linear quadratic regulator control algorithm, and output a preliminary control strategy; Based on the movement data, predict the future movement data of the stirring head through the predictive control algorithm, and adjust the control input based on the prediction result to optimize the preliminary control strategy; Generate a control signal based on the optimized control strategy, and send the control signal to the servo motor of the slide table to control the servo motor; Use the difference between the real-time movement data of the stirring head and the prediction result as feedback data, and adjust the control strategy parameters to adapt to different welding working conditions.

[0006] Preferably, the dynamic model of the stirring head includes: ; Wherein, is the mass of the slide table; is the acceleration of the slide table; is the damping coefficient; is the speed of the sliding table; is the elastic coefficient; is the position of the sliding table; is the feedback force, which is a control force signal generated based on the difference between the moving data and the prediction result.

[0007] Preferably, the cost function of the optimal linear quadratic regulator control algorithm is used to minimize the weighted sum of the position error and the speed error, and the cost function is: ; where, is the position vector of the sliding table; is the speed vector of the sliding table; and are positive definite matrices, representing the penalty degrees for controlling the position error and the speed error; is the cost function, and the goal is to minimize this cost function to obtain the optimal control input.

[0008] Preferably, the quadratic regulator control algorithm generates the optimal control input based on the dynamic model and adjusts the speed and position of the sliding table through the servo motor to control the movement of the stirring head.

[0009] Preferably, the objective function of the predictive control algorithm is: ; where, is the future time step predicted by the system the position of the stirring head; is the time step the target trajectory; is the time step the control input at; and are the weight matrices, representing the relative importance of the state and the control input; is the transpose operator; is the prediction step.

[0010] Preferably, when optimizing the control input, the predictive control algorithm is based on the maximum speed and acceleration of the sliding table and adjusts the preliminary control strategy in real time to minimize the control error at future moments.

[0011] Preferably, the moving data further includes real-time detection of the material and thickness of the workpiece, and the dynamic model of the stirring head is updated in real time according to the detection results.

[0012] Preferably, the acquisition of the movement data of the friction welding machine stirring head further includes, after acquiring the movement data, preprocessing the movement data, and the preprocessing includes filtering the movement data to eliminate sensor noise, removing outliers, and performing coordinate system unification and time alignment processing on the movement data.

[0013] Preferably, the adjustment of the control strategy parameters to adapt to different welding conditions specifically includes: Calculating the error between the real-time movement data of the stirring head and the prediction result, And dynamically adjusting the weight matrices in the optimal linear quadratic regulation control algorithm and the predictive control algorithm based on this error.

[0014] A friction welding machine slide control device, comprising: An acquisition module, which based on a sensor, acquires the movement data of the friction welding machine stirring head in real time and establishes a dynamic model of the stirring head; An optimal control calculation module, which based on the movement data and the dynamic model, uses the optimal linear quadratic regulation control algorithm to calculate the optimal control input and outputs a preliminary control strategy; A predictive control module, which is used to predict the future movement data of the stirring head based on the movement data through the predictive control algorithm, and adjusts the control input based on the prediction result to optimize the preliminary control strategy; A control signal generation module, which is used to generate a control signal based on the optimized control strategy and send the control signal to the servo motor of the slide to drive the servo motor to control the movement of the slide; A feedback regulation module, which is used to compare the difference between the real-time movement data of the stirring head and the prediction result, and adjusts the control strategy parameters according to the difference to adapt to different welding conditions.

[0015] In summary, the present invention includes at least one of the following beneficial technical effects: 1. By using the difference between the real-time movement data of the stirring head and the prediction result as feedback data to adjust the control strategy of the servo motor of the slide, the present invention forms a closed-loop control mechanism, which can continuously monitor the actual operating state of the stirring head and dynamically adjust the control parameters, forming a continuous and high-speed feedback loop. This technical solution realizes a significant improvement in the position control accuracy of the stirring head, and finally effectively improves the weld uniformity and overall quality. Compared with traditional open-loop or low-sensitivity closed-loop systems, this solution breaks through the problem of lag in the response of the control system. Especially in high-speed welding tasks, it can better ensure the high consistency of the trajectory and solve the problem of weld position deviation.

[0016] 2. The present invention dynamically estimates the position of the stirring head at future moments by introducing a predictive control algorithm, which can greatly improve the stability and control accuracy of the system, effectively suppress trajectory drift, and thus solve the problem in the prior art that due to the fact that the control system can only passively respond to the current state, control errors are likely to occur at the welding speed change or the path turning point.

[0017] 3. The present invention establishes a stirring head dynamics model based on the material and thickness of the workpiece, and the control strategy weight can be self-adjusted according to the error between the actual and predicted results to dynamically correct the dynamics model, and this technical solution greatly enhances the adaptability of the system to complex welding conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a schematic flow chart of the method of the present invention; Figure 2 is a schematic diagram of the device system architecture of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0019] The following will further elaborate on the present invention in conjunction with the attached Figure 1 for a more detailed description.

[0020] The present invention provides a method for controlling a friction welding machine slide, including the following steps: Collect the movement data of the stirring head of the friction welding machine in real time, where the movement data includes the movement direction and movement speed, and establish a stirring head dynamics model; In this embodiment, to ensure the accuracy and dynamic adaptability of the subsequent control strategy generation, it is necessary to accurately perceive the real-time movement state of the stirring head and establish an accurate dynamics model on this basis. Based on this, a preprocessing module centered on data collection and modeling is introduced at the initial stage of the method execution. This module not only provides basic physical modeling support for the control algorithm, but also significantly improves the control accuracy and the robustness of the system response through data cleaning and structural optimization.

[0021] Specifically, first collect the movement data of the stirring head in real time through the sensor system installed on the friction welding machine slide. The movement data includes at least two dimensions of information: one is the movement direction of the stirring head during welding, and the other is the movement speed. The sensor can be a laser displacement meter, an optical encoder or an inertial measurement unit (IMU) to ensure the high-resolution and low-latency performance of the collected data.

[0022] Generally, the original sensing data will be affected by factors such as environmental noise, welding heat interference, and equipment vibration. Therefore, the present invention introduces a preprocessing mechanism. Specifically: First, collect the movement data of the stirring head during the welding process through multiple high-precision sensors set on the sliding table or the stirring head. This data includes two basic dimensions: the movement direction and the movement speed. Among them, the movement direction can be measured by a three-axis accelerometer and an angular velocity sensor, and the movement speed can be obtained by a laser interferometer or a linear grating scale.

[0023] In a possible implementation, before the collected raw data is input into the system model, a series of preprocessing operations are required to enhance its stability and credibility. Generally, this preprocessing process includes but is not limited to the following steps: Perform filtering processing on the sensor data, such as using techniques like band-pass filters and Kalman filters to effectively eliminate the high-frequency fluctuation components introduced by sensor noise or environmental interference; Implement an outlier rejection mechanism. For example, within a certain time window, use sliding mean and standard deviation analysis means to identify and remove data points that deviate from the statistical characteristics; Perform coordinate system unification and time alignment operations to synchronize multi-source data under a unified spatial and time reference to ensure the consistency and timeliness of the modeling data.

[0024] Specifically, after the data undergoes the above preprocessing, it will enter the dynamic modeling stage of the stirring head. In the present invention, the following dynamic model is used to model the movement process of the sliding table.

[0025] As an option, in some embodiments, a material perception module is further introduced. This module real-time detects the material parameters and thickness information of the workpiece to be welded, and thereby dynamically updates the constructed dynamic model of the stirring head. In a possible implementation, use ultrasonic sensors or electromagnetic induction sensing devices to complete the identification of the material type (such as aluminum alloy, stainless steel) and thickness (such as 3mm, 5mm), and feedback the detection results to the model parameter calculation module in real time.

[0026] Specifically, based on the movement data after the above preprocessing, the present invention establishes the following dynamic model of the stirring head: ; Wherein, is the mass of the sliding table; is the acceleration of the sliding table; is the damping coefficient; is the speed of the sliding table; is the elastic coefficient; is the position of the sliding table; is the feedback force, which is a control force signal generated according to the difference between the movement data and the prediction result.

[0027] In some embodiments, the kinetic model provides a theoretical basis for the calculation of subsequent optimal control inputs. Through the above model, a control strategy can be derived based on control algorithms (such as linear quadratic regulation control or model predictive control), so as to precisely control the driving of the slide servo motor and achieve the coordinated regulation of the position and speed of the stirring head.

[0028] It should be noted that in practical applications, the parameters of the kinetic model can be obtained through experimental identification or online adaptive update methods to adapt to the effects of different welding working conditions, equipment states, or material changes.

[0029] In summary, this embodiment not only starts from mobile data collection and processing to construct a systematic data support mechanism, but also combines physical modeling methods to form a complete slide control modeling link, laying a solid technical foundation for the generation of subsequent optimal control strategies. Based on mobile data and the kinetic model, calculate the optimal control input through the optimal linear quadratic regulation control algorithm and output the preliminary control strategy; In this embodiment, after completing the mobile data collection and preprocessing of the friction stir welding head and constructing a kinetic model that conforms to the actual working conditions, in order to effectively regulate the movement of the slide, the present invention further introduces the optimal linear quadratic regulation control algorithm (Linear - Quadratic - Regulator, LQR) to generate a preliminary control strategy for the current system state. This control algorithm plays a core role in connecting physical modeling and control execution in this method. The control input calculated by it not only has stability and real - time performance, but also has the basic function that can be further iteratively optimized by the predictive control algorithm.

[0030] Specifically, first, according to the slide kinetic model obtained in the previous steps, for the convenience of optimal control design, convert the above - mentioned second - order differential equation into the standard state - space form. Generally, define the system state vector as: ; where, is the position of the slide; is the speed of the slide; The control input is defined as the servo - motor control signal , representing the force actually applied to the slide (or motor command). Thus, the standard linear system model is obtained: ; where, the system matrix and the input matrix According to the , , It is obtained from numerical derivation, and its specific structure can be automatically calculated by an engineering recognition or simulation system.

[0031] To achieve optimal control of the sliding table motion state, in this embodiment, based on minimizing the weighted sum of the position error and the velocity error of the following cost function, the cost function is: ; Wherein, is the position vector of the sliding table; is the velocity vector of the sliding table; and are positive definite matrices, representing the penalty degrees for controlling the position error and the velocity error; is the cost function, and the goal is to minimize this cost function to obtain the optimal control input.

[0032] This cost function aims to minimize the weighted sum of the position and velocity errors within the entire time domain while balancing the energy consumption of the control input; Specifically, if it is set that: ; ; Wherein, represents the control weight for the position error; represents the control weight for the velocity error; represents the suppression intensity of the control signal energy. These parameters can be adjusted according to the system response characteristics.

[0033] Furthermore, by solving the Algebraic-Riccati-Equation, the optimal feedback gain matrix can be obtained, and then a linear feedback control law is constructed: ; In this embodiment, the calculated control input is output as a preliminary control strategy and sent to the servo motor via the control signal generation module to drive the sliding table to move, thereby realizing the indirect control of the stirring head.

[0034] As an option, in some specific working conditions, this control strategy can integrate hardware constraints such as the maximum speed and maximum acceleration of the sliding table, and by adjusting , the weight parameters of the matrix, the control input can not only meet the performance index requirements but also take into account the system execution ability.

[0035] Specifically, in a possible implementation, the system can automatically adjust the weight coefficients according to dynamic working conditions such as workpiece thickness and welding speed requirements, thereby improving the adaptability of the control strategy. For example, when welding thick workpieces, appropriately increase , to strengthen the suppression of position error; while in high-speed welding of thin workpieces, can be increased to strengthen the control of speed response.

[0036] Therefore, this control strategy not only has good robustness and engineering feasibility, but also provides high-quality initial control input values for the subsequent predictive control and feedback regulation modules, thereby ensuring the coordination and high-precision operation of the entire sliding table control system.

[0037] Based on the moving data, predict the future moving data of the stirring head through the predictive control algorithm, and adjust the control input based on the prediction results to optimize the preliminary control strategy; In this embodiment, after the preliminary control strategy is generated, considering the uncertainty and dynamic changes of the working conditions during the welding process, such as material thermal expansion, friction coefficient fluctuations, and structural transient responses, etc., relying solely on the feedback control input calculated based on the current state may be difficult to meet the higher-precision trajectory control requirements. In order to improve the prediction ability of the control system for future states and further optimize the control performance, the present invention introduces a predictive control algorithm into the control system. This algorithm is based on the currently collected moving data, combined with the established dynamic model, to estimate the future motion state of the stirring head in advance, and accordingly correct the preliminary control strategy, thereby significantly improving the continuity, stability of the welding trajectory and the forward-looking of the control response.

[0038] In this embodiment, model predictive control (MPC) framework is adopted for predictive control. Its core lies in: using a sliding time window to predict the state evolution trend of the system in the future period of time, and on the basis of the prediction, determining the current control input by solving an optimization problem, so that the future state is as close as possible to the target trajectory.

[0039] Generally, the predictive control module takes the moving data at the current moment as input, and the data includes the current position, current speed, welding path direction of the stirring head, and the target trajectory point set. Deduce the future state through the state space model, and use the following objective function (cost-function) as the optimization criterion: ; Among them, is the future time step predicted by the system the position of the stirring head; is the time step the target trajectory; is the time step control input; and is the weight matrix, representing the relative importance of the state and control input; is the transpose operator; is the prediction step.

[0040] In a specific implementation, model predictive control adopts a recursive approach to gradually estimate the state variable and uses the state update at each step as the initial condition for the next step to realize the progressive construction of the prediction chain. All control inputs within the prediction window are obtained by unified solution of an optimization solver (such as a QP solver).

[0041] Specifically, model predictive control not only considers the trajectory error but also introduces the physical operation limits of the sliding table. For example, in some embodiments, to prevent the sliding table from having too fast acceleration or deceleration or over-limit behavior, the system adds the following constraint conditions: Maximum speed of the sliding table Constraint: ; Maximum acceleration constraint of the sliding table: ; Maximum control input amplitude constraint; ; The above constraints are introduced into the optimization model in the form of linear inequalities to ensure the engineering feasibility and equipment safety of the optimization results.

[0042] In a possible implementation, the model predictive control module outputs and applies only the first control input in each control cycle to the current moment, and the remaining prediction results are discarded. This strategy called "receding-horizon-control" ensures that the system is optimized based on the latest state at each moment, improving the dynamic response ability.

[0043] As an option, in some embodiments, if there is a significant deviation between the prediction result and the current preliminary control strategy, the system can perform weighted adjustment on it according to the preset fusion weight. For example, the final control input can be expressed as: ; where is the preliminary control strategy previously obtained by the LQR algorithm; is the control input optimized by the model predictive control module; is the weighting coefficient, and the value range is , which can be dynamically adjusted according to the error size.

[0044] Through this control input fusion mechanism, the present invention realizes dual regulation based on feedback and prediction, improves the robustness and disturbance resistance of the system, and is particularly suitable for high-quality welding control tasks under complex or dynamically changing working conditions.

[0045] Moreover, to improve the accuracy of the prediction model, the present invention also supports online correction of model parameters during the control process. For example, the system matrix or gain term in the state space model is continuously adjusted through the deviation between the actual feedback data and the predicted value to achieve model adaptive adjustment.

[0046] In summary, the predictive control method proposed by the present invention not only makes full use of current mobile data and dynamic models, but also integrates target trajectories, hardware constraints, and multi-strategy fusion mechanisms, constituting a control optimization module with a complete structure, diverse strategies, and accurate prediction, providing accurate, controllable, and future-oriented control path generation capabilities for the friction welding machine slide system.

[0047] Generate a control signal based on the optimized control strategy and send a control signal to the servo motor of the slide to control the servo motor; In this embodiment, after completing the optimized adjustment of the control input based on the predictive control algorithm, to achieve accurate tracking of the target trajectory by the stirring head, the present invention further introduces a control signal generation and distribution mechanism, which is used to convert the optimized control input into an electrical control signal executable by the servo drive system and send it to the servo motor in the slide control system, thereby driving the slide to complete the corresponding displacement response. This process belongs to the final execution link in the system control chain and has a close logical and functional coupling relationship with previous modules such as preliminary control strategy generation, predictive control correction, and real-time system state feedback, forming a complete closed-loop control process from state perception, control calculation to execution feedback.

[0048] Specifically, the optimized control input is generally represented in the form of a digital signal, and its physical meaning usually corresponds to the speed setting value, acceleration adjustment amount, or displacement increment command in the moving direction of the slide. After being processed by a digital controller (such as a DSP or an embedded main control unit), this input quantity is converted into a corresponding analog control voltage signal or pulse width modulation (PWM) signal for the servo driver to receive.

[0049] Generally, there is a control signal scheduling module inside the controller, which is used to output updated control instructions within each control period according to the real-time sampling period. In the controller, this module receives the control input output by the predictive control algorithm and generates an analog quantity signal through a digital-to-analog conversion interface (such as a DAC module). This signal is input to the servo driver after filtering, buffering, and amplitude adjustment.

[0050] Specifically, after receiving a control signal, the servo driver applies a corresponding current to the motor according to the control mode set inside it (such as speed control, position control, or current control) to achieve closed-loop control of torque or speed, and then drives the mechanical slide to move. In this embodiment, the control mode is preferably a position-speed composite control mode, where: The upper-layer master control system sets the target displacement or trajectory; The control input is provided by the MPC optimization module; The lower-layer servo controller realizes real-time response adjustment of the actual motor position and speed.

[0051] In a possible implementation, to ensure the response consistency and time synchronization of control instructions, the system implements a fixed-frequency interrupt scheduling mechanism in each control cycle to ensure that the control signal can complete the entire process of acquisition, calculation, conversion, and distribution within the sampling period.

[0052] As an option, in some embodiments, to prevent the motor excitation from being unstable or overloaded due to sudden changes in the control signal, a signal smoothing module is also configured in the controller to perform slow-varying processing on the control signal through methods such as interpolation filtering, moving average, or lag compensation, thereby avoiding mechanical shock to the actuator.

[0053] Specifically, in the signal generation module, the sliding filter algorithm can perform weighted averaging on the control quantity of each cycle and the value of the previous cycle For example, set the weight coefficient such that the actually issued signal is: ; where, , to adjust the dependence of the current instruction on the previous state, thereby controlling the smoothness of signal changes.

[0054] At the system execution level, after being amplified by the driver, the control signal is applied to the servo motor winding in the form of a constant voltage or pulsed current, causing electromagnetic excitation and prompting the rotor to move. The rotor converts the rotation into linear motion through a lead screw, slide rail, or linear module, and finally realizes two-dimensional or three-dimensional spatial movement control of the stirring head. This movement process is monitored in real time by a position encoder, and the encoder feedback signal is sampled and enters the upper-layer control system for constructing a closed-loop adjustment path.

[0055] In some embodiments, to improve the anti-interference and stability of the control system, the control signal generation module also introduces mechanisms such as voltage amplitude limitation, slope limitation, and signal interruption protection. For example, in the high-speed trajectory segment, to prevent oscillations caused by frequent large changes in the input control signal, the system can set a maximum change rate threshold to constrain the change amplitude of each cycle signal.

[0056] In addition, to adapt to different types of slide table execution units, the present invention also supports a control signal reconstruction mechanism based on device parameters. During the initialization phase, the system reads characteristic data such as the servo motor model, electrical parameters, and torque constant, automatically calculates the mapping relationship between the control signal and the actual displacement, and dynamically matches the signal range, offset, and calibration factor to ensure the accuracy and device adaptability of the control signal output.

[0057] In summary, by constructing a complete control signal generation and execution mechanism, the present invention realizes seamless docking from optimizing the control input to physical drive execution. It not only improves the execution consistency and dynamic response performance of the system, but also provides stable and reliable final-stage control guarantee for high-precision trajectory control. As the terminal output interface of the entire control process, this link plays a key role in achieving accurate movement of the stirring head and precise tracking of the welding path; Taking the difference between the real-time movement data of the stirring head and the prediction result as feedback data, adjust the control strategy parameters to adapt to different welding conditions.

[0058] In this embodiment, to achieve adaptive adjustment for different welding conditions, the system uses the error between the real-time movement data of the stirring head and the prediction result as a feedback signal, thereby adjusting the parameters of the control strategy. This adjustment process dynamically adjusts the key parameters in the control algorithm by calculating the error in real time and according to the change of the error, especially the weight matrix in the optimal linear quadratic regulator (LQR) algorithm and the predictive control algorithm, to achieve the best control effect and welding quality.

[0059] Specifically, first, the real-time movement data of the stirring head is collected and compared with the ideal result predicted by the system to calculate the error between the two. This error reflects the deviation between the actual situation and the theoretical expectation in the current welding process, and is usually calculated by the following formula: ; where, represents the error, is the actual position data of the real-time stirring head, is the ideal position data calculated by the prediction model.

[0060] Based on this error, the system will dynamically adjust the weight matrices in the LQR and predictive control algorithms to optimize the control effect. Specifically, the control matrix in the LQR algorithm and the optimization problem in the predictive control algorithm will be updated according to the real-time error feedback. The weight matrices of the LQR controller usually include the state weighting matrix and the control input weighting matrix , and these matrices will be dynamically adjusted according to the error feedback. The optimization goal of the control strategy is to ensure the stability and efficiency of the welding process while minimizing the system error.

[0061] In a possible implementation, the specific process of dynamically adjusting the weight matrix of the LQR control algorithm can be expressed as: ; where, is the optimized control gain matrix, is the error, is the control input, and are positive definite matrices, and at each moment , the control matrix will be adjusted according to the current error, is the upper limit of the optimization time interval, representing the time range of control strategy optimization.

[0062] In addition, the weight matrix of the predictive control algorithm will also be dynamically adjusted according to the magnitude of the error. For example, the system may update the weight matrix in the predictive control through the following formula: ; where, is the adjusted weight matrix of the predictive control, is the original weight matrix of the predictive control, is the adjustment factor, represents the error. The selection of the adjustment factor will be optimized according to the changes in the actual welding working conditions to better adapt to the welding requirements under different working conditions.

[0063] Therefore, through the above adjustment process, the system can adapt to different welding working conditions in real time, ensure the best control effect during the welding process, and thus improve the welding quality and efficiency. Generally, with the changes in the welding environment or parameters, the system can adjust the control parameters according to the real-time feedback to ensure the stability of the process and the consistency of the welding effect. As an option, the feedback control system can also combine other sensor data to further improve the adaptability to complex welding working conditions.

[0064] Moreover, the process of adjusting the control strategy not only depends on error calculation but can also incorporate other auxiliary information, such as parameters like welding current, voltage, temperature, etc., to further optimize the control strategy and ensure that the welding quality meets the expected standards.

[0065] Furthermore, to further improve the efficiency and intelligence level of the system in repetitive welding tasks, the system designs a data storage and reuse mechanism. During each welding process, various data including the position information of the stirring head, welding temperature, current, voltage, welding speed, welding depth, weld quality evaluation indicators, etc. can be collected in real-time and stored in the database. The data not only covers the workpiece geometric information and welding parameter set values but also includes the system's records of the dynamic responses during the welding process, thus forming a detailed welding condition dataset.

[0066] When the system detects that the workpiece to be welded currently has similar or identical geometric features and material properties to a certain type of workpiece in the historical records, the system can quickly retrieve and extract the corresponding process parameters and movement path data based on these historical welding data. The path data includes the three-dimensional trajectory executed by the stirring head during the welding process, the speed change curves of each section of the path, attitude change information, etc.; the welding parameters include key control parameters such as rotational speed, feeding speed, axial pressure, preheating temperature, etc. This parameter setting method based on historical experience significantly reduces the test verification time required before welding repetitive workpieces and improves the work efficiency and welding quality consistency of the system.

[0067] Even further, to achieve the prediction of the movement trend of the stirring head during the welding process, a machine learning algorithm, especially the Support-Vector-Machine (SVM) model, is introduced in this embodiment. This model can use the welding process data of historical similar workpieces as training samples to learn the dynamic behavior patterns of the stirring head under specific working conditions. By constructing the optimal hyperplane in the high-dimensional feature space, SVM can effectively identify the mapping relationship between input features (such as the current welding process, position information, speed, historical error, etc.) and the movement state of the stirring head.

[0068] In practical applications, the system first matches the type of the workpiece to be welded currently, the working condition information with the historical database, selects a set of historical data closest to the current task as the training set, trains the support vector machine model to predict the movement trend of the stirring head in the next period of time, including its displacement, speed change, and attitude adjustment direction, etc. The prediction results can be used as an important reference input for the control system and, together with the real-time error feedback mechanism, further enhance the response ability and stability of the control algorithm to the welding dynamic process.

[0069] Through the above mechanism, the system not only realizes the effective utilization of historical welding experience, but also introduces an intelligent prediction function on this basis, providing higher adaptive ability and process optimization support for complex welding environments. Especially in mass production and standardized production scenarios, this method greatly reduces the need for initial process tests, effectively reduces the operator's dependence on parameter setting, and improves the standardization and replicability of welding processes.

[0070] Combined with the attached Figure 2 , this embodiment also provides a method for controlling the slide table of a friction welding machine, including: An acquisition module that collects the movement data of the friction welding machine's stirring head in real time based on sensors and establishes a dynamic model of the stirring head; An optimal control calculation module that calculates the optimal control input using the optimal linear quadratic regulation control algorithm based on the movement data and the dynamic model, and outputs a preliminary control strategy; A predictive control module that predicts the future movement data of the stirring head through a predictive control algorithm based on the movement data, and adjusts the control input based on the prediction results to optimize the preliminary control strategy; A control signal generation module that generates a control signal based on the optimized control strategy and sends the control signal to the servo motor of the slide table to drive the servo motor to control the movement of the slide table; A feedback adjustment module that compares the difference between the real-time movement data of the stirring head and the prediction results, and adjusts the control strategy parameters according to the difference to adapt to different welding working conditions.

[0071] In this embodiment, the acquisition module is responsible for collecting the movement data of the stirring head during the welding process in real time, mainly including parameter information such as movement direction, speed, and acceleration. To achieve high-precision data acquisition, the system can be equipped with various types of sensors, such as laser rangefinders, optical encoders, inertial measurement units (IMUs), etc. The raw data collected will undergo preprocessing steps such as filtering, outlier removal, and time alignment before entering the control system to improve data quality and stability. In addition, this module also has a material identification function, which can automatically adjust the model parameters according to changes in the workpiece material or thickness, providing a more practical modeling basis for subsequent control algorithms.

[0072] Based on the real-time data provided by the acquisition module and the dynamic modeling results of the stirring head, the optimal control calculation module uses the optimal linear quadratic regulation control algorithm (LQR) to calculate the optimal control input at the current moment. This algorithm aims to minimize the weighted sum of the stirring head position error and speed error, ensuring a good balance between control strategy accuracy and energy efficiency. The output preliminary control strategy not only responds quickly but also has strong robustness, which can be used as the basic input for the predictive control algorithm and provide a reference for the next step of optimization.

[0073] The predictive control module plays a role in forward-looking adjustment in the control logic. Based on the current system state and modeling results, this module uses the model predictive control algorithm (MPC) to simulate and predict the motion state of the stirring head in the next several time steps, and optimizes the control input by combining the system target trajectory and physical constraints (such as maximum speed and acceleration). In each control cycle, the system only applies the first input value in the optimal sequence, and the rest is replaced with the update of the new cycle, so as to achieve rolling optimization. This module significantly enhances the stability and accuracy of the system under complex paths or dynamic working conditions.

[0074] The control signal generation module converts the optimized control input from the previous step into an electrical control signal to control the servo motor to perform corresponding actions. Signal conversion can adopt methods such as digital-to-analog conversion, PWM pulse width modulation, or voltage-current mapping to ensure high compatibility and low-latency response between the control signal and the servo driver. To prevent the impact of excitation mutations on the equipment, the system can also add a signal smoothing mechanism to limit the change rate of the input signal, thereby improving the stability of control execution and the service life of the mechanical structure.

[0075] The feedback adjustment module provides closed-loop feedback capabilities for the entire control process. By comparing the error between the actual motion state of the stirring head and the predicted value, the system can evaluate the effectiveness of the current control strategy in real time and fine-tune the control parameters, especially for adaptive optimization of the weight matrix in the LQR and MPC algorithms. This mechanism allows the system to automatically adjust the control sensitivity and response intensity according to different welding working conditions (such as changes in material hardness and complexity of weld structure), improving the overall adaptability and control stability of the system.

[0076] In summary, the method proposed in this embodiment realizes a complete closed-loop control process from data perception, intelligent decision-making to precise execution. The coordinated cooperation of multiple modules not only improves the control accuracy and path consistency of the stirring head during welding, but also enhances the real-time adaptability of the system to changing working conditions, greatly promoting the development of friction welding equipment towards intelligence and high performance.

[0077] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A control method for the sliding table of a friction welding machine, characterized in that, It includes the following steps: Collect the moving data of the friction stir welding head in real time. The moving data includes the moving direction and moving speed, and establish a dynamic model of the stir welding head; Based on the moving data and the dynamic model, calculate the optimal control input through the optimal linear quadratic regulator (LQR) control algorithm, and output a preliminary control strategy; Based on the moving data, predict the future moving data of the stir welding head through the model predictive control (MPC) algorithm, and adjust the control input based on the prediction result to optimize the preliminary control strategy; Generate a control signal based on the optimized control strategy, and send the control signal to the servo motor of the slide to control the servo motor; Use the difference between the real-time moving data of the stir welding head and the prediction result as feedback data, and adjust the control strategy parameters to adapt to different welding conditions.

2. The method for controlling a sliding table of a friction welding machine according to claim 1, wherein The dynamic model of the stir welding head includes: ； Among them, is the mass of the sliding table; is the acceleration of the sliding table; is the damping coefficient; is the velocity of the sliding table; is the elastic coefficient; is the position of the sliding table; is the feedback force, which is a control force signal generated according to the difference between the moving data and the prediction result.

3. A method for controlling a slide table of a friction welding machine according to claim 2, characterized in that, The cost function of the optimal linear quadratic regulator control algorithm is used to minimize the weighted sum of the position error and the speed error. The cost function is: ； wherein, is the position vector of the sliding table; is the velocity vector of the sliding table; and are positive definite matrices, representing the penalty degrees for controlling the position error and the velocity error; is the cost function, and the goal is to minimize this cost function to obtain the optimal control input.

4. A method for controlling a sliding table of a friction welding machine according to claim 3, characterized in that, The quadratic regulator control algorithm generates the optimal control input based on the dynamic model, and adjusts the speed and position of the slide through the servo motor to control the movement of the stir welding head.

5. A method for controlling a sliding table of a friction welding machine according to claim 1, characterized in that, The objective function of the model predictive control algorithm is: ； Among them, is the future time step predicted by the system position of the stirring head; is the time step target trajectory; is the time step control input of; and is the weight matrix, representing the relative importance of the state and the control input; is the transpose operator; is the prediction step number.

6. The method for controlling a slide table of a friction welding machine according to claim 5, characterized in that, When optimizing the control input, the model predictive control algorithm is based on the maximum speed and acceleration of the slide, and adjusts the preliminary control strategy in real time to minimize the control error at future moments.

7. A method for controlling a slide table of a friction welding machine according to claim 3, characterized in that, The moving data also includes real-time detection of the material and thickness of the workpiece, and updates the dynamic model of the stir welding head in real time according to the detection results.

8. A method for controlling a sliding table of a friction welding machine according to claim 1, characterized in that The collection of the moving data of the friction stir welding head further includes, after collecting the moving data, preprocessing the moving data. The preprocessing includes filtering the moving data to eliminate sensor noise, removing outliers, and performing coordinate system unification and time alignment processing on the moving data.

9. A method for controlling a sliding table of a friction welding machine according to claim 1, characterized in that, The adjustment of the control strategy parameters to adapt to different welding conditions specifically includes: Calculate the error between the real-time moving data of the stir welding head and the prediction result, And dynamically adjust the weight matrix in the optimal linear quadratic regulator control algorithm and the model predictive control algorithm based on this error.

10. A control device for the slide table of a friction welding machine, based on a method for controlling the slide table of a friction welding machine according to any one of claims 1-9, characterized in that, It includes: A collection module, which collects the moving data of the friction stir welding head in real time based on sensors and establishes a dynamic model of the stir welding head; An optimal control calculation module, which calculates the optimal control input using the optimal linear quadratic regulator control algorithm based on the moving data and the dynamic model, and outputs a preliminary control strategy; A model predictive control module, which is used to predict the future moving data of the stir welding head through the model predictive control algorithm based on the moving data, and adjust the control input based on the prediction result to optimize the preliminary control strategy; A control signal generation module, which is used to generate a control signal based on the optimized control strategy and send the control signal to the servo motor of the slide to drive the servo motor to control the movement of the slide; A feedback regulation module, which is used to compare the difference between the real-time moving data of the stir welding head and the prediction result, and adjust the control strategy parameters according to the difference to adapt to different welding conditions.

Images