A method and system for detecting pressure on a shot peening machine.

Abstract

This invention belongs to the field of pressure detection technology, specifically relating to a pressure detection method and system for shot peening machines. The method includes: synchronously acquiring fluid data from a valve assembly and motion data from a robot, and aligning them spatiotemporally; constructing a pipeline geometric torsion characteristic model based on the real-time angle values ​​of the robot's joints and the change in end-effector position, and calculating the pipeline geometric torsion at each moment; calculating the total pressure attenuation value of the flexible pipeline using a rheological geometric dynamics model based on the gas mass flow rate, pipeline geometric torsion, and the rate of change of pipeline geometric torsion; reconstructing the estimated pressure at the nozzle end based on the reference pressure at the valve assembly outlet and the total pressure attenuation value, and combining this with the predicted pressure attenuation value for future moments based on robot trajectory planning data to generate control commands sent to the electro-proportional valve. This invention combines robot trajectory prediction with feedforward control, achieving high-precision constant pressure control under complex dynamic operations.

Description

Technical Field

[0001] This invention relates to the field of pressure detection technology. More specifically, this invention relates to a method and system for detecting pressure on a shot peening machine. Background Technology

[0002] Key components of aero-engines typically have thin walls, low rigidity, and complex curved surfaces. To improve their fatigue resistance, they must undergo high-precision shot peening. Dual-robot shot peening machines are typical high-end equipment in this field, using two six-axis industrial robots to hold the nozzle and workpiece respectively, achieving full-coverage processing through multi-axis linkage.

[0003] The core pneumatic architecture of existing high-end shot peening control systems typically includes a precision electro-proportional valve, a high-flow pneumatic pressure reducing valve, and a high-precision pressure transmitter. The control logic generally adopts a single closed-loop feedback mode, that is, the pressure sensor is installed at the valve group outlet or pressure stabilizing chamber, and the controller adjusts the opening of the proportional valve according to the feedback value of the sensor to maintain a constant downstream pressure.

[0004] However, in actual aerospace precision machining scenarios, in order to ensure the robot's flexible movement, the valve assembly and nozzle are usually connected by a flexible wear-resistant pipeline several meters long. This causes the pressure sensor to only read the value at the valve assembly end and cannot sense the actual state at the end of the flexible pipeline. At the same time, when machining the air intake edge, exhaust edge, or blade root radius of the blade, the robot arm needs to make large-angle twisting and folding movements. As a result, the flexible pipeline undergoes severe bending or even local flattening, causing the pressure loss along the pipeline to increase exponentially and drastically, resulting in a significant decrease in the actual pressure ejected from the nozzle.

[0005] Furthermore, when the robot performs rapid posture changes, the rapid expansion and contraction of the flexible pipeline will change the pipeline volume, generating airflow suction or compression, resulting in millisecond-level drastic fluctuations in end pressure. Existing technology cannot solve the problem of inconsistency between the pressure detection value and the actual operating value caused by the dynamic deformation of the pipeline, which seriously restricts the processing consistency of weak rigid parts in aerospace. Summary of the Invention

[0006] To address the technical problems of the aforementioned sensors being unable to sense the actual pressure at the end of flexible pipelines and being unable to cope with pressure fluctuations caused by dynamic deformation of the pipelines, this invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a pressure detection method for a shot peening machine, comprising: synchronously acquiring fluid data at the valve assembly end and motion data of a robot, and resampling the motion data based on the fluid sampling time to achieve spatiotemporal alignment; constructing a pipeline geometric torsion characteristic model based on the real-time angle values ​​of the robot joints and the change in end position, and calculating the pipeline geometric torsion at each moment; calculating the total pressure attenuation value of the flexible pipeline through a rheological geometric dynamics model based on the gas mass flow rate, pipeline geometric torsion, and the rate of change of pipeline geometric torsion; reconstructing the estimated pressure at the nozzle end based on the reference pressure at the valve assembly outlet and the total pressure attenuation value, and combining the predicted pressure attenuation value for future moments based on robot trajectory planning data to generate a control command sent to an electro-proportional valve.

[0008] This invention system constructs a mapping model between pose and flow resistance, transforming the robot's spatial motion state into the geometric torsion characteristics of the pipeline in real time. Combining fluid dynamics principles, it accurately estimates the pressure attenuation caused by steady-state bending and dynamic deformation of the pipeline, effectively overcoming the physical limitation that sensors cannot be installed at the end of the motion. Simultaneously, by combining trajectory look-ahead technology to predict future pressure change trends, it drives the electro-proportional valve to implement feedforward compensation before the actual occurrence of interference. This predictive control strategy effectively eliminates the physical lag of pneumatic transmission and the response delay of the actuator, suppressing air pressure pulsations caused by the robot's rapid acceleration and deceleration. Thus, it achieves high-fidelity constant pressure control with full attitude and zero delay in complex working conditions of dual-robot collaborative operation, ensuring the consistency of shot peening quality in key curved areas of weakly rigid aerospace components.

[0009] Preferably, the formula for calculating the geometric tortuosity of the pipeline is: ;in, Indicates time Pipeline geometric torsion; Indicates the first The flow resistance sensitivity weight of each joint is given, with the wrist joint having a greater weight than the basic joints. Indicates time No. Real-time angle values ​​of each joint; This indicates that the flexible pipeline is in a naturally straightened, stress-free state. Reference angles corresponding to each joint; Indicates the first The maximum physical travel of each joint; Indicates the bending nonlinearity index; This represents the tensile coupling coefficient, which is obtained by performing a tensile resistance calibration experiment. Indicates time Change in the position of the robot's end effector; This indicates the maximum allowable elastic expansion and contraction length of the flexible pipeline.

[0010] This invention constructs a refined mathematical model of pipeline geometric distortion. By introducing the flow resistance sensitivity weights of each joint, the bending nonlinearity index, and the stretching coupling coefficient, the physical distortion degree of the flexible pipeline under nonlinear large-angle bending and axial stretching conditions is determined. This model can truly reflect the exponential growth law of pipeline flow resistance with attitude change, thereby significantly improving the accuracy of describing the geometric characteristics of pipeline under complex spatial attitudes.

[0011] Preferably, the formula for calculating the estimated total pressure loss of the flexible pipeline is: ;in, Indicates time The estimated total pressure loss of the flexible pipeline; Indicates the basic flow resistance coefficient; Indicates time The gas mass flow rate; Indicates the geometric impedance gain factor; Indicates time Pipeline geometric torsion; Indicates time Pipeline deformation rate; Indicates time The reference pressure at the valve assembly outlet; This represents the dynamic damping coefficient.

[0012] To address the problem that existing static flow resistance models cannot capture transient pressure fluctuations caused by rapid pipeline deformation, this invention constructs a total pressure loss estimation model that includes a steady-state friction term and a dynamic damping term. The steady-state term is based on Darcy's formula and has been geometrically modified to calculate the increase in secondary flow resistance caused by bending, while the dynamic term uses the pipeline deformation rate and dynamic damping coefficient to capture pressure fluctuations caused by rapid changes in pipeline volume, thereby achieving accurate estimation of pressure loss across the entire frequency band.

[0013] Preferably, the step of reconstructing the estimated pressure at the nozzle tip based on the reference pressure at the valve assembly outlet and the total pressure attenuation value includes: based on the current time... The reference pressure at the valve assembly outlet Compared with the calculated total pressure attenuation value The estimated pressure at the nozzle tip is reconstructed through interpolation. .

[0014] Preferably, the calculation formula sent is: ;in, Indicates time Control commands sent to the electric proportional valve; Indicates the proportional gain coefficient; This indicates the target shot peening pressure set in the process parameters. Indicates time The estimated pressure at the nozzle tip obtained from the reconstruction; Indicates the integral gain coefficient; Indicates time Pressure deviation, , Indicates time The estimated pressure at the nozzle tip obtained from the reconstruction; Indicates the feedforward compensation gain; This represents the pressure attenuation value obtained based on predictions of the robot's future trajectory.

[0015] This invention introduces a feedforward compensation term based on future trajectory prediction on the basis of traditional PID feedback control. The control algorithm uses the predicted pressure decay value to adjust the opening of the electro-proportional valve in advance, and applies compensation before the actual increase in physical resistance or pressure fluctuation occurs. This overcomes the pure lag characteristic of the pneumatic system, effectively eliminates pressure drop and oscillation during rapid robot movement, and achieves high-fidelity constant pressure control.

[0016] Preferably, the method for obtaining the predicted pressure attenuation value for future moments based on robot trajectory planning data includes: using the trajectory planning data of the robot controller to read the future moments in advance. Motion data, future moments The unit is ; Future moments Substituting the motion data into the pipeline geometric torsion characteristic model, the future time is calculated. The pipe geometry distortion; using current fluid data and combining it with future data. The geometric torsion of the pipeline is used to predict future moments. Total pressure loss estimate for flexible pipelines , which is the pressure attenuation value obtained based on the prediction of the robot's future trajectory.

[0017] In response to the physical phenomenon that abrupt changes in fluid state usually lag behind mechanical motion, this invention utilizes known trajectory planning data in the robot controller to read the motion posture at future moments in advance and substitute it into the model calculation. This strategy of trading space for time solves the problem that relying solely on current fluid data cannot cope with future load changes, enabling the control system to predict flow resistance in advance based on the changing trend of pipeline geometry, greatly improving the dynamic response capability of the system.

[0018] Preferably, the basic flow resistance coefficient is obtained by controlling the robot arm to extend and collecting flow rate and pressure drop data under different pressure levels, and then fitting the data using the least squares method.

[0019] Preferably, the dynamic damping coefficient is obtained by controlling the robot to swing back and forth rapidly at different frequencies, monitoring the correlation between the pressure fluctuation amplitude and the pipeline deformation rate, and solving the dynamic damping coefficient through time-domain response analysis.

[0020] Preferably, the bending nonlinearity index and geometric impedance gain factor are obtained during the bending resistance calibration experiment performed during the system initialization phase: keeping the robot stationary, controlling the electro-proportional valve to output a constant gas mass flow rate; controlling the robot end effector to gradually increase the joint angle according to a preset trajectory, recording the real-time angle values ​​of the joints and the actual measured pressure drop under different bending postures; for the constructed joint regression equation, using the least squares method, with the real-time angle value of the joint as input and the actual pressure drop as the target, performing nonlinear regression analysis on the joint regression equation, thereby simultaneously calculating the best-matching bending nonlinearity index and geometric impedance gain factor.

[0021] In a second aspect, the present invention provides a shot peening machine pressure detection system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the aforementioned shot peening machine pressure detection method is implemented.

[0022] By adopting the above technical solution, a computer program for the shot peening machine pressure detection method is generated and stored in a memory for loading and execution by a processor. This allows for the creation of a terminal device based on the memory and processor, making it convenient to use.

[0023] The beneficial effects of this invention are as follows:

[0024] This invention system constructs a mapping model between pose and flow resistance, transforming the robot's spatial motion state into the geometric torsion characteristics of the pipeline in real time. Combining fluid dynamics principles, it accurately estimates the pressure attenuation caused by steady-state bending and dynamic deformation of the pipeline, effectively overcoming the physical limitation that sensors cannot be installed at the end of the motion. Simultaneously, by combining trajectory look-ahead technology to predict future pressure change trends, it drives the electro-proportional valve to implement feedforward compensation before the actual occurrence of interference. This predictive control strategy effectively eliminates the physical lag of pneumatic transmission and the response delay of the actuator, suppressing air pressure pulsations caused by the robot's rapid acceleration and deceleration. Thus, it achieves high-fidelity constant pressure control with full attitude and zero delay in complex working conditions of dual-robot collaborative operation, ensuring the consistency of shot peening quality in key curved areas of weakly rigid aerospace components. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating a shot peening machine pressure detection method according to the present invention;

[0026] Figure 2This is a schematic diagram illustrating the time-domain changes in the robot's motion state and the geometric characteristics of the pipeline;

[0027] Figure 3 This is a schematic diagram illustrating the comparison between the pressure decay value calculated based on the rheological geometry dynamics model and the actual physical pressure drop;

[0028] Figure 4 This is a schematic diagram illustrating the comparison between the present invention and the prior art in terms of the actual pressure response at the nozzle tip. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0031] This invention discloses a method for detecting pressure on a shot peening machine tool, referring to... Figure 1 This includes steps S1-S4:

[0032] S1: Synchronously acquire fluid data and robot motion data at the valve assembly end, and resample the motion data based on the fluid sampling time to achieve spatiotemporal alignment.

[0033] It should be noted that, since the fluid control system of the shot peening machine and the robot motion control system operate in different clock domains, the sampling frequency of fluid data is much higher than that of robot motion data, and there is a significant difference in the data transmission delay between the two. This causes the physical state to be misaligned on the time axis when data fusion is performed directly, which in turn leads to incorrect compensation calculations. Therefore, this embodiment establishes a dual-channel data acquisition mechanism and uses a high-order interpolation algorithm to resample low-frequency data in order to achieve strict alignment of heterogeneous data at the microsecond-level time accuracy.

[0034] Specifically, the system establishes two parallel high-speed data acquisition channels: the fluid sensing channel reads the pressure sensor values ​​and gas mass flow rate count values ​​installed at the valve group outlet in real time through a high-frequency analog signal acquisition card to obtain the reference pressure at the valve group outlet. and gas mass flow rate The sampling frequency is 1kHz; the spatial sensing channel reads the real-time angle values ​​of the robot's six-axis joints via industrial real-time Ethernet. , , , , and The sampling frequency is 100Hz.

[0035] Furthermore, the system uses fluid sampling time Based on this, a cubic spline interpolation algorithm is used to resample the low-frequency robot joint angle data to construct a physical state vector synchronized with the fluid data. .

[0036] S2: Based on the real-time angle values ​​of the robot's joints and the change in end-effector position, construct a pipeline geometric torsion feature model and calculate the pipeline geometric torsion at each moment.

[0037] It should be noted that, since the flexible pipeline is physically attached to the surface of the robotic arm, the contribution of the rotation of different joints of the robot to the degree of pipeline bending is quite different. Moreover, the flow resistance caused by the cross-sectional deformation of the pipeline under extreme postures exhibits a nonlinear exponential growth characteristic, and a simple linear superposition of angles cannot accurately describe this physical phenomenon. Therefore, this embodiment introduces a weighting coefficient and normalization processing mechanism based on physical experiment calibration to construct a geometric twist index that can quantify the overall distortion degree of the pipeline, mapping the complex six-dimensional spatial posture into a single flow resistance characteristic scalar.

[0038] First, joint sensitivity analysis and weight allocation are performed: The robot's six-axis joints include three basic joints and three wrist joints. The basic joints are mainly responsible for large-scale spatial positioning. Their movement usually only changes the overall direction of the pipeline, causing less local bending, and are therefore assigned low weights. The wrist joints are responsible for fine adjustment of the nozzle posture. Their movement range is large and directly causes sharp bends at the end of the pipeline, having the greatest impact on flow resistance, and are therefore assigned high weights. Specifically, the flow resistance sensitivity weight of each joint is determined through a single-axis excitation experiment, that is, the robot is controlled to rotate each joint one by one and the pressure drop change is measured. The pressure drop change of all joints is normalized to obtain the flow resistance sensitivity weight of each joint.

[0039] Furthermore, based on the real-time angle values ​​of the robot's joints and the change in end-effector position, a pipeline geometric torsion feature model is constructed to calculate the pipeline geometric torsion. This feature is the core characteristic for measuring the increasing trend of pipeline air resistance; the formula for calculating this index is:

[0040]

[0041] in, Indicates time The geometric tortuosity of the pipeline; the larger the value, the higher the degree of bending and distortion of the flexible pipeline. Indicates the first The flow resistance sensitivity weight of each joint is used to characterize the difference in the effect of different joint movements on pipeline bending. The weight of the wrist joint is usually greater than that of the basic joints. Indicates time No. Real-time angle values ​​of each joint; This indicates that the flexible pipeline is in a naturally straightened, stress-free state. Reference angles corresponding to each joint; Indicates the first The maximum physical travel of each joint is used to normalize the angle deviation. It represents the bending nonlinearity index, used to fit the physical law that the flow resistance of a pipeline increases exponentially with the increase of the bending angle; This represents the tensile coupling coefficient, which characterizes the effect of axial tension on flow resistance in a pipeline. It is obtained by performing a tensile resistance calibration experiment. Indicates time The change in robot end-effector position refers to the change in the distance between the robot end-effector position and the fixed point in the pipeline. This indicates the maximum allowable elastic expansion and contraction length of the flexible pipeline, used to normalize the amount of stretching.

[0042] The process of obtaining the tensile coupling coefficient by performing a tensile resistance calibration experiment is as follows: keep the robot in a straight position, control only the linear axis movement of the robot to stretch the flexible pipeline, record the pressure drop changes under different stretching lengths, and calculate the slope of the linear relationship between the stretching length and the pressure drop growth rate. This slope is the tensile coupling coefficient.

[0043] It should be noted that, with the robot's joint angles Deviating from the natural state The degree of increase, or the amount of pipe stretching Increase the geometric tortuosity of the pipeline It will show a significant non-linear increase, and high-weighted joint movements will lead to Faster growth, thus accurately reflecting the changing trend of fluid resistance inside flexible pipelines.

[0044] For example, a schematic diagram of the time-domain changes in the robot's motion state and the pipeline's geometric features is shown below. Figure 2 As shown; the curve corresponding to the geometric torsion of the pipeline shows a trapezoidal change in amplitude as the robot joints move, reflecting the entire process of the flexible pipeline from stretching to bending and then straightening: during the 2-4 second period, the torsion gradually increases as the robot bends its arm, maintains a large angle of bending for 4-6 seconds, and quickly straightens and returns to its original position for 6-7 seconds; the curve corresponding to the pipeline deformation rate shows the speed of change in the pipeline geometry, where a sharp negative peak appears during the rapid straightening at 6-7 seconds, i.e., the instant of action switching, which characterizes the dynamic disturbance source caused by the sudden change in pipeline volume.

[0045] S3: Calculate the total pressure attenuation value of the flexible pipeline using a rheological geometric dynamics model based on the gas mass flow rate, pipeline geometric tortuosity, and the rate of change of pipeline geometric tortuosity.

[0046] It should be noted that since the pressure loss of gas-solid two-phase flow in flexible pipelines includes not only the steady-state frictional resistance caused by pipeline bending, but also the transient pressure pulsation caused by the volume change due to rapid pipeline deformation, the static flow resistance model cannot capture this dynamic disturbance. Therefore, this embodiment combines fluid mechanics principles, introduces the basic flow resistance coefficient, geometric impedance gain factor, and dynamic damping coefficient, and constructs a pressure attenuation estimation model that includes steady-state friction terms and dynamic damping terms to achieve accurate prediction of pressure fluctuations across the entire frequency band.

[0047] Specifically, based on the rheological geometric dynamics model, the estimated total pressure loss of the flexible pipeline is calculated; the formula for this index is:

[0048]

[0049] in, Indicates time The estimated total pressure loss of the flexible pipeline; The basic flow resistance coefficient represents the inherent frictional characteristics of a flexible pipeline in a straight state. This coefficient is obtained by controlling the robot arm to extend and collecting flow and pressure drop data under different pressure levels, and then fitting the data using the least squares method. Indicates time The gas mass flow rate; The geometric impedance gain factor represents the amplification factor of the flow resistance coefficient per unit geometric twist. This factor is obtained through regression analysis of multi-posture compound motion experiments. Indicates time Pipeline geometric torsion; Indicates time The pipeline deformation rate was obtained by processing the pipeline geometric torsion and time using the five-point difference method. Indicates time The reference pressure at the valve assembly outlet; The dynamic damping coefficient represents the intensity of the air pressure suction or compression effect caused by the rapid deformation of the flexible pipeline volume. This coefficient is obtained by controlling the robot to swing back and forth rapidly at different frequencies and monitoring the correlation between the pressure fluctuation amplitude and the pipeline deformation rate, and then solving the dynamic damping coefficient through time-domain response analysis.

[0050] Among them, the bending nonlinearity index in the pipeline geometric torsion characteristic model And the geometric impedance gain factor in the formula for estimating the total pressure loss of flexible pipelines. The bending resistance calibration experiment was performed during the system initialization phase. The specific process was as follows: keeping the robot stationary, the electro-proportional valve was controlled to output a constant gas mass flow rate. The robot's end effector was controlled to gradually increase the joint angles according to a preset trajectory, and the real-time angle values ​​of the joints under different bending postures and the actual measured pressure drop were recorded; the constructed joint regression equation was then analyzed. Using the least squares method, the real-time angle values ​​of the joints are... For input, actual voltage drop To achieve this, nonlinear regression analysis is performed on the joint regression equation to simultaneously calculate the best-matching bending nonlinear index. and geometric impedance gain factor .

[0051] It should be noted that the total pressure decay value It consists of two parts, the first part The steady-state friction term describes the change in flow rate. Increase or pipe twist As the frictional resistance increases, the steady-state frictional resistance increases significantly. Based on the friction resistance, following Darcy's formula, This is a geometric correction term, reflecting the increase in secondary flow resistance caused by bending; the latter part... The dynamic damping term describes the effect of rapid deformation in the flexible pipeline. When the value is large, additional transient pressure fluctuations will occur. When the robot bends the pipe rapidly, the positive value calculated by this item represents the instantaneous increase in pressure, that is, positive deformation leads to increased pressure. Conversely, negative deformation leads to decreased pressure. This allows the system to not only know the degree of bending of the flexible pipe, but also to sense the bending speed of the flexible pipe, thereby achieving accurate quantification of the dynamic process.

[0052] For example, a schematic diagram comparing the pressure attenuation value calculated based on the rheo-geodynamic model with the actual physical pressure drop is shown below. Figure 3 As shown, the two exhibit a high degree of overlap across the entire time domain, and the blue dashed line contains the simulation model error and random noise. Therefore, the model constructed in this invention can accurately capture the pressure loss caused by the steady-state bending and dynamic deformation of the flexible pipeline, providing a reliable data foundation for feedforward compensation.

[0053] S4: Reconstruct the estimated pressure at the nozzle end based on the reference pressure at the valve assembly outlet and the total pressure decay value, and combine it with the predicted pressure decay value for future moments based on robot trajectory planning data to generate control commands to be sent to the electro-proportional valve.

[0054] It should be noted that, due to the inherent physical lag in compressed air transmission and the response delay of the electric proportional valve adjustment action, traditional real-time feedback control is difficult to eliminate pressure fluctuations caused by rapid robot movements within milliseconds. Therefore, this embodiment utilizes the trajectory planning data of the robot controller to obtain the future motion posture in advance, combines it with the aforementioned model to predict the future pressure attenuation, and applies feedforward compensation before the actual occurrence of the disturbance, thereby overcoming the pure lag characteristic of the system.

[0055] First, based on the current time The reference pressure at the valve assembly outlet Compared with the calculated estimated total pressure loss of the flexible pipeline The estimated pressure at the nozzle tip is reconstructed through interpolation. This value is recorded as the actual process pressure; the system regards this value as the actual pressure acting on the surface of the part, and uses it for process monitoring and quality traceability, filling the blind spot of the end without physical sensors.

[0056] Subsequently, based on feedforward control with trajectory look-ahead, the trajectory planning data of the robot controller is used to read future moments in advance. The motion data, namely the real-time angle values ​​of all joints and the position of the robot's end effector, in future moments. The unit is ; Future moments Substituting the motion data into the pipeline geometric torsion characteristic model, the future time is calculated. The system assumes that fluid state abrupt changes lag behind mechanical motion within a small prediction time domain. Therefore, it utilizes current fluid data, i.e., the current gas mass flow rate and the reference pressure at the valve assembly outlet, combined with future pipeline geometry parameters, to obtain the future... The geometric torsion of the pipeline is used to predict future moments. Total pressure loss estimate for flexible pipelines As the pressure attenuation value obtained based on the prediction of the robot's future trajectory, this processing method effectively solves the problem that fluid data at future moments cannot be known in advance, and realizes the prediction of flow resistance changes in advance by utilizing the trend of geometric configuration changes.

[0057] Finally, based on the estimated pressure at the nozzle tip obtained from the reconstruction and the predicted pressure decay value at future moments, a control command is generated and sent to the electro-proportional valve. The specific calculation formula is as follows:

[0058]

[0059] in, Indicates time Control commands sent to the electric proportional valve; This indicates the target shot peening pressure set in the process parameters. Indicates time The estimated pressure at the nozzle tip obtained from the reconstruction; Indicates time Pressure deviation, , Indicates time The estimated pressure at the nozzle tip obtained from the reconstruction; This represents the proportional gain coefficient, used to quickly eliminate the current error; The integral gain coefficient is used to eliminate steady-state cumulative error. The proportional gain coefficient and integral gain coefficient are determined through step response experiments. In closed-loop control mode, a pressure step signal is given to the system, the pressure response curve is observed, and the parameters are tuned according to the Ziegler-Nichols rule to make the system reach the critical damping state. This represents the feedforward compensation gain, used to adjust the intervention intensity of the feedforward quantity. In this embodiment... Setting it to 1.0 means full compensation is performed. In other embodiments, implementers can choose to set it between 0.8 and 1.2 according to the system stability requirements. The pressure decay value is obtained based on the prediction of the robot's future trajectory.

[0060] Among them, control commands A feedforward compensation term is added to the conventional PID feedback regulation. When the algorithm predicts that the robot is about to perform an action that will increase resistance, the feedforward compensation term... It will increase immediately before the actual pressure drop occurs, making... The pressure rises, thereby driving the electric proportional valve to open the valve port in advance; when the actual action occurs and the pipeline resistance truly increases, that is, at the instant when the physical resistance actually increases, the pressure drop is just offset. The increased pressure after the valve is raised just offsets the increased resistance, so that the pressure at the end of the nozzle remains unchanged, achieving constant pressure control with zero delay.

[0061] For example, a schematic diagram showing the comparison between the present invention and the prior art in terms of the actual pressure response at the nozzle tip is shown below. Figure 4 As shown; the process target pressure is constant at 0.4 MPa; the curve corresponding to the prior art shows a severe pressure drop during the robot's bending action and significant oscillations during rapid movements, indicating a severe under-spraying state; for the curve corresponding to the present invention, since the system predicts the pressure drop in advance and automatically increases the valve output, through feedforward compensation, it closely follows the process target line throughout the process, only showing extremely small fluctuations during violent movements, proving that the present invention can effectively eliminate the interference of flexible pipeline deformation and achieve high-fidelity constant pressure control.

[0062] This invention also discloses a shot peening machine pressure detection system, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a shot peening machine pressure detection method according to the present invention.

[0063] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A method for detecting pressure on a shot peening machine, characterized in that, include: Simultaneously collect fluid data from the valve assembly and motion data from the robot, and resample the motion data based on the fluid sampling time to achieve spatiotemporal alignment; Based on the real-time angle values ​​of the robot's joints and the change in end-effector position, a characteristic model of the pipeline's geometric torsion is constructed to calculate the pipeline's geometric torsion at each moment, satisfying the following: ； Indicates time Pipeline geometric torsion; Indicates the first The flow resistance sensitivity weight of each joint is given, with the wrist joint having a greater weight than the basic joints. Indicates time No. Real-time angle values ​​of each joint; This indicates that the flexible pipeline is in a naturally straightened, stress-free state. Reference angles corresponding to each joint; Indicates the first The maximum physical travel of each joint; Indicates the bending nonlinearity index; This represents the tensile coupling coefficient, which is obtained by performing a tensile resistance calibration experiment. Indicates time Change in the position of the robot's end effector; Indicates the maximum allowable elastic expansion and contraction length of the flexible pipeline; The total pressure attenuation of the flexible pipeline is calculated using a rheo-geodynamic model based on the gas mass flow rate, pipeline geometric tortuosity, and the rate of change of pipeline geometric tortuosity. The estimated total pressure loss of the flexible pipeline meets the following requirements: ； Indicates time The estimated total pressure loss of the flexible pipeline; Indicates the basic flow resistance coefficient; Indicates time The gas mass flow rate; Indicates the geometric impedance gain factor; Indicates time Pipeline geometric torsion; Indicates time Pipeline deformation rate; Indicates time The reference pressure at the valve assembly outlet; Indicates the dynamic damping coefficient; The estimated pressure at the nozzle tip is reconstructed based on the reference pressure at the valve assembly outlet and the total pressure decay value, including: According to the current time The reference pressure at the valve assembly outlet Compared with the calculated total pressure attenuation value The estimated pressure at the nozzle tip is reconstructed through interpolation. ; Combined with the predicted pressure decay values ​​for future moments based on robot trajectory planning data, control commands are generated and sent to the electric proportional valve.

2. The method for detecting pressure on a shot peening machine tool according to claim 1, characterized in that, The formula for calculating the control command sent to the electro-proportional valve is: ； in, Indicates time Control commands sent to the electric proportional valve; Indicates the proportional gain coefficient; This indicates the target shot peening pressure set in the process parameters. Indicates time The estimated pressure at the nozzle tip obtained from the reconstruction; Indicates the integral gain coefficient; Indicates time Pressure deviation, , Indicates time The estimated pressure at the nozzle tip obtained from the reconstruction; Indicates the feedforward compensation gain; This represents the pressure attenuation value obtained based on predictions of the robot's future trajectory.

3. The method for detecting pressure on a shot peening machine tool according to claim 2, characterized in that, The method for obtaining the predicted pressure decay value for future moments based on robot trajectory planning data includes: By using trajectory planning data from the robot controller, future moments can be read in advance. Motion data, future moments The unit is ; Future moments Substituting the motion data into the pipeline geometric torsion characteristic model, the future time is calculated. The pipe geometry distortion; using current fluid data and combining it with future data. The geometric torsion of the pipeline is used to predict future moments. Total pressure loss estimate for flexible pipelines , which is the pressure attenuation value obtained based on the prediction of the robot's future trajectory.

4. The method for detecting pressure on a shot peening machine tool according to claim 1, characterized in that, The basic flow resistance coefficient is obtained by controlling the robot arm to extend and collecting flow rate and pressure drop data under different pressure levels, and then fitting the data using the least squares method.

5. The method for detecting pressure on a shot peening machine tool according to claim 1, characterized in that, The dynamic damping coefficient is obtained by controlling the robot to swing back and forth rapidly at different frequencies and monitoring the correlation between the pressure fluctuation amplitude and the pipeline deformation rate, and then solving the dynamic damping coefficient through time-domain response analysis.

6. The method for detecting pressure on a shot peening machine tool according to claim 1, characterized in that, The bending nonlinearity index and geometric impedance gain factor were obtained during the bending resistance calibration experiment performed during the system initialization phase: keeping the robot stationary, controlling the electro-proportional valve to output a constant gas mass flow rate; controlling the robot end effector to gradually increase the joint angle according to a preset trajectory, recording the real-time angle values ​​of the joints and the actual measured pressure drop under different bending postures; for the constructed joint regression equation, using the least squares method, with the real-time angle value of the joint as input and the actual pressure drop as the target, performing nonlinear regression analysis on the joint regression equation, thereby simultaneously calculating the best-matching bending nonlinearity index and geometric impedance gain factor.

7. A pressure detection system for a shot peening machine tool, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a shot peening machine pressure detection method according to any one of claims 1-6.

Images