A mechanical arm plasma torch integrated control method and system

Abstract

The application discloses a kind of mechanical arm plasma torch integrated control method and system, it is related to plasma spraying technical field, in the system operation, establish motion characteristic and process parameter mapping model;Obtain workpiece three-dimensional information to generate job path point and parameter set;Real-time acquisition end pose and torch state parameter;Based on mapping relationship dynamic adjustment motion compensation and process execution variable;Through multidimensional sensing feedback realizes closed-loop control and coating growth state prediction.The system includes central integrated control module, multi-axis mechanical arm driving module, plasma process execution module and sensing perception module.The application realizes the deep integration control of mechanical arm and torch by the above-mentioned scheme, eliminates the coating defect at trajectory switching place, significantly improves coating uniformity, bonding strength and the stability of complex component processing.

Description

Technical Field

[0001] This invention relates to the field of plasma spraying technology, specifically to an integrated control method and system for a robotic arm plasma spray gun. Background Technology

[0002] With the widespread application of advanced manufacturing technologies, the deep integration of industrial robots and thermal spraying processes has become a core driving force in the field of modern surface engineering. In industries such as aerospace, energy equipment, and transportation, the use of automation to strengthen and protect material surfaces can significantly improve the wear resistance, high-temperature resistance, and oxidation resistance of key components, which has significant application value for ensuring the safe operation of heavy equipment and extending its service life.

[0003] Among them, the integrated control technology of the robotic arm plasma spray gun is the core link to achieve precision coating processing. It aims to ensure that the spray gun jet can always uniformly and efficiently cover the workpiece surface according to the preset physical characteristics under diverse processing paths by deeply coordinating the motion trajectory and posture accuracy of the multi-axis robotic arm with the process parameters of the plasma spraying system.

[0004] Traditional robotic arm plasma spray gun control technology suffers from several drawbacks: First, the robotic arm motion control system and the plasma spraying execution system lack real-time deep integration, often operating asynchronously or loosely. This makes it difficult to precisely match the spraying start / stop control with the robotic arm's acceleration / deceleration, easily leading to coating buildup or missed spraying at workpiece trajectory switching points. Second, the system lacks dynamic adaptability to complex process parameters. When dealing with irregular curved surfaces, it struggles to adjust the plasma arc power and powder delivery rate in real time based on instantaneous shifts in spraying distance and angle, thus affecting the coating's bonding strength and composition. The first problem is the lack of dense fabrication; the second is the lack of closed-loop real-time feedback from multi-dimensional sensor data in the control system, making it difficult to effectively predict and intervene in environmental disturbances and hardware wear during the spraying process, resulting in poor consistency and stability in long-term operation; finally, the lack of unified software and hardware interface standards leads to long system debugging cycles and high system reconstruction costs, making it difficult to meet the requirements of flexible production for rapid production changeover. These problems collectively limit the efficiency and quality of high-quality thermal spraying. Therefore, developing a method and system that can achieve full-dimensional deep integration control of robotic arms and plasma spray guns has become an urgent technical challenge to be solved in the field of industrial automation. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an integrated control method and system for a robotic arm plasma spray gun, solving the problems mentioned in the background section.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an integrated control method for a robotic arm plasma spray gun, comprising the following steps: S1. Establish a nonlinear mapping model between the motion characteristics of a multi-axis robotic arm and the parameters of plasma spraying process; S2. Obtain the three-dimensional geometric information of the workpiece to be processed, and generate the set of working path points of the robotic arm and the initial process parameter set of the spray gun accordingly. S3. During the operation, the position and posture data of the robotic arm end and the instantaneous operating status parameters of the plasma spray gun are collected in real time. S4. Based on the mapping relationship between motion characteristics and process parameters, dynamically calculate and adjust the motion speed compensation of the robotic arm and the execution variables of the plasma spraying system; S5. Achieve closed-loop control of the spraying process through multi-dimensional sensor feedback, and output the prediction results of the coating growth state.

[0007] Preferably, step S1 specifically includes the following steps: S11. Obtain the end effector jitter characteristics and trajectory deviation distribution of a multi-axis robotic arm under different load and acceleration conditions through dynamic modeling; S12. Establish a physical characteristic model of the plasma spray gun, including the influence of arc voltage, arc current, main gas flow rate and powder feeding rate on the velocity and temperature of the sprayed particles. S13. Through feature extraction algorithms, the end-effector linear velocity, angular velocity, and attitude angle change rate of the robotic arm are correlated with the deposition efficiency and coating thickness distribution of plasma spraying to form a multi-dimensional collaborative control matrix.

[0008] Preferably, step S2 specifically includes the following steps: S21. Use a 3D vision sensor to acquire point cloud data of the workpiece surface, and perform noise reduction and surface reconstruction. S22. Based on the preset coating thickness index and overlap rate requirements, plan equally spaced scanning paths on the reconstructed curved surface and determine the normal vector information of the path points. S23. Based on the curvature change law of the path points, preset the target movement speed of the robotic arm and the rated power of the plasma spray gun at each path point to form a process path guidance document.

[0009] Preferably, step S3 specifically includes the following steps: S31. Obtain the encoder feedback values ​​of each joint and the real-time spatial coordinates of the end effector through the high-speed data interface of the robotic arm controller at a preset sampling frequency. S32. Obtain the real-time output parameters of the plasma power supply, the cooling water temperature of the heat exchange system, and the instantaneous speed of the powder feeder through the industrial fieldbus. S33. Synchronously record the spraying distance offset and spraying angle deviation detected by the sensor, and align all data on a unified time scale.

[0010] Preferably, step S4 specifically includes the following steps: S41. Calculate the spatial deviation between the current actual trajectory and the preset path, and solve the compensation torque of each joint of the robotic arm by using the inverse kinematics algorithm; S42. Based on the fluctuation of instantaneous spraying distance, the arc power of the plasma is corrected in real time using a mapping model to maintain a constant energy density when particles hit the workpiece surface. S43. Based on the change in linear velocity of the robotic arm during acceleration and deceleration, synchronously adjust the powder feeder flow rate to ensure that the amount of material deposited per unit length of the path remains consistent.

[0011] Preferably, the nonlinear mapping model adopts a deep residual network structure, in which the input layer receives the motion state vector of the robotic arm and the environmental disturbance factor, the hidden layer extracts feature associations through multi-layer nonlinear transformation, and the output layer provides suggested values ​​for process parameter correction.

[0012] Preferably, the dynamic modeling process considers the joint friction torque, centripetal torque, and Coriolis torque of the multi-axis robotic arm, and establishes a complete dynamic state-space expression through the Lagrange equation to predict trajectory tracking errors under high dynamic motion.

[0013] Preferably, when generating the set of working path points for the robotic arm, a specific path smoothing algorithm is used for the edge area and abrupt curved surface of the workpiece. Under the premise of ensuring the perpendicularity of the spraying angle, the impact vibration of the robotic arm is reduced by means of circular arc transition or spline curve interpolation.

[0014] Preferably, the pose data includes real-time values ​​of the three translational degrees of freedom and three rotational degrees of freedom of the robotic arm end effector in the Cartesian coordinate system, and the data update frequency is synchronized with the robotic arm interpolation cycle.

[0015] Preferably, the instantaneous operating parameters of the plasma spray gun also include the pressure drop at both ends of the nozzle and the pressure fluctuation of the powder-feeding carrier gas, which are used to monitor the stability of the plasma jet and serve as a reference for adjusting process parameters.

[0016] Preferably, when dynamically calculating the motion speed compensation of the robotic arm, a predictive control algorithm is introduced to predict the trajectory trend of multiple future sampling periods based on the current motion state. The optimal control increment is solved by minimizing the objective function, thereby suppressing the impact of motion lag on the coating quality.

[0017] Preferably, the execution variable correction of the plasma spraying system includes real-time fine-tuning of the ratio of main gas to auxiliary gas to change the compression degree of the plasma arc, thereby achieving dynamic control of the spraying beam diameter to adapt to path requirements of different widths.

[0018] Preferably, the multi-dimensional sensing feedback in step S5 includes using an infrared thermal imaging sensor to monitor the instantaneous temperature field distribution on the workpiece surface, and automatically triggering a protection command to accelerate the robotic arm or reduce the arc power when the local temperature exceeds a preset threshold.

[0019] Preferably, the prediction result of the coating growth state is based on numerical simulation algorithm and real-time acquired process data to calculate the cumulative amount of the current layer thickness and predict the coating density and bonding strength level after the completion of the entire operation process.

[0020] Preferably, the present invention also provides an integrated control system for a robotic arm plasma spray gun, used to implement the above-mentioned method, comprising: a central integrated control module for executing core algorithm logic and task scheduling; a multi-axis robotic arm drive module, the input end of which is connected to the motion command output end of the central integrated control module for precisely executing the motion trajectory; a plasma process execution module, including a plasma power supply, a gas path control cabinet and a powder feeder, which is controlled by the process adjustment signal output by the central integrated control module; and a sensing module, including a high-precision encoder, a vision sensor, a pressure sensor and a temperature sensor, for real-time monitoring of the system's full-dimensional operating status and transmitting the signals back to the central integrated control module.

[0021] Preferably, the central integrated control module has a built-in high-speed logic processor and a large-capacity storage unit, supports multi-protocol data parsing, and can realize transparent forwarding and deep integration of the robotic arm control protocol and the plasma device control protocol.

[0022] Preferably, the multi-axis robotic arm drive module has a hardware-level synchronization enable interface to ensure that the robotic arm can start moving along the trajectory with a certain time delay at the same time as receiving the spraying start command.

[0023] Preferably, the plasma process execution module adopts a modular design, and the subsystems communicate with each other through a real-time deterministic network to ensure that the adjustment response time of process parameters is in the microsecond range.

[0024] Preferably, the sensing module uses a specific data fusion algorithm to perform weight allocation and noise filtering on information from different physical dimensions, thereby improving the reliability and accuracy of the feedback data.

[0025] Preferably, the central integrated control module also has fault diagnosis and safety prediction functions, which can issue early warnings and take degraded operation strategies before potential faults occur by analyzing historical data trends.

[0026] This invention provides an integrated control method and system for a robotic arm plasma spray gun, which has the following beneficial effects: (1) During system operation, this invention realizes deep integrated control of the robotic arm and the spraying system from the physical layer to the logical layer by constructing a nonlinear mapping model of the robotic arm's motion characteristics and plasma spraying process parameters. First, this deep integration effectively solves the problem of asynchronous motion and process in traditional control. By directly introducing the real-time motion state of the robotic arm into the calculation loop of the process parameters, it ensures that the spraying energy and material supply can fluctuate precisely with the change of motion speed during complex start-stop, turning and variable acceleration processes. This fundamentally eliminates the phenomenon of coating accumulation or missed spraying at trajectory switching points and significantly improves the uniformity of coating thickness and processing accuracy.

[0027] (2) This invention acquires the three-dimensional geometric information of the workpiece to be processed and performs dynamic parameter calculations by combining the curvature changes of the path points, thus endowing the system with extremely strong adaptive capabilities. When facing workpieces with complex curved surfaces such as aero-engine blades and large irregular structural parts, the system can correct the negative impacts caused by spraying distance offset and angle deviation in real time. By dynamically adjusting the arc power, powder feeding rate and airflow ratio, the system ensures the constant physical properties of the plasma jet under variable operating environments, thereby significantly enhancing the bonding strength between the coating and the substrate and improving the density of the microstructure, meeting the stringent requirements of high-end equipment for high-performance coatings.

[0028] (3) By introducing a closed-loop real-time feedback mechanism based on multi-dimensional sensor data, this invention constructs an intelligent control system with self-correction capabilities. Real-time monitored data such as end-effector pose, arc state, and ambient temperature are used to compensate for prediction errors and environmental disturbances in real time. This not only overcomes the long-term impact of hardware wear on process stability but also significantly improves the consistency of coating quality under continuous large-scale operations. Simultaneously, the coating growth prediction results generated based on real-time data provide a scientific basis for online quality monitoring of the production process, reducing subsequent inspection costs and scrap rates.

[0029] (4) The integrated control system structure proposed in this invention coordinates the operation of multiple modules through a unified central integrated control module, solving the problem of system reconfiguration difficulties caused by inconsistent software and hardware interface standards. This highly integrated and standardized design significantly shortens the debugging cycle of new products, reduces the complexity and cost of system maintenance and upgrades, and can quickly respond to the needs of flexible production. By using a deep residual network to construct a mapping model, the system's ability to handle multivariable and strongly coupled problems is enhanced, making the suggestions for process parameter correction more scientific and forward-looking. The use of inverse kinematics algorithm to solve the compensation torque and predictive control algorithm to optimize the motion trajectory greatly improves the trajectory tracking accuracy of the multi-axis robotic arm under high dynamic response, providing a stable motion base for precision spraying. Through infrared thermal imaging sensing monitoring and temperature field closed-loop control, the risk of thermal deformation or coating cracking of workpieces during thermal spraying is effectively prevented, ensuring the structural integrity of complex components. The system's fault diagnosis and safety prediction functions, through long-term analysis of full-dimensional sensor data, realize the transformation from passive maintenance to proactive prevention, greatly improving the system's reliability and equipment utilization. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the overall technical solution architecture of the integrated control method for the robotic arm plasma spray gun proposed in this invention. Figure 2 This is a schematic diagram of the core principle framework of the nonlinear mapping model between the motion characteristics of the robotic arm and the parameters of the plasma spraying process in this invention; Figure 3 This is a logical flowchart of the closed-loop control and dynamic compensation of the spraying process based on multi-dimensional sensor feedback in this invention. Detailed Implementation

[0031] 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 embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0032] Example 1 This invention provides an integrated control method for a robotic arm plasma spray gun. Please refer to [link / reference]. Figure 1 This embodiment provides an integrated control method for a robotic arm plasma spray gun, with its core application scenario being the spraying of ceramic coatings on the surface of high-pressure turbine blades for aero-engines. Due to the extremely complex irregular curved surfaces and variable curvature characteristics of the blades, extremely high requirements are placed on the accuracy of the robotic arm's motion trajectory and the real-time coupling of plasma process parameters during the spraying process.

[0033] In step S1, a nonlinear mapping model between the motion characteristics of the multi-axis robotic arm and the plasma spraying process parameters is first established. This process begins with in-depth dynamic modeling of a six-axis industrial robotic arm. Dynamic modeling is used to obtain the end effector jitter characteristics and trajectory deviation distribution of the multi-axis robotic arm under different loads and acceleration conditions. In specific engineering implementations, the system identifies the inertial response characteristics of the robotic arm when performing highly dynamic actions by collecting current fluctuations at each joint during high-speed operation and small positional deviations fed back by the encoder. This process fully considers the mutual coupling between joints, including the centripetal force effect caused by rotation and the Coriolis force interference caused by the simultaneous movement of each axis. Through refined modeling of the frictional torque between joints, especially the parameterized definition of the influence of lubrication state and ambient temperature on the switching process between static and dynamic friction, a physical model that reflects the true dynamic response of the robotic arm's end effector in three-dimensional space is constructed.

[0034] Simultaneously, a physical characteristic model of the plasma spray gun was established. This model delves into the energy conversion efficiency and jet dynamics during the plasma arc generation process. Specifically, the system meticulously records and analyzes the influence of arc voltage and arc current fluctuations on the degree of plasma ionization, thereby determining their decisive role in the acquisition of kinetic and thermal energy by the sprayed particles. By precisely defining the proportional relationship between the main gas flow rate (e.g., argon or nitrogen) and the auxiliary gas flow rate (e.g., hydrogen), the physical envelope of the plasma jet formed after these gases are heated and expanded by the arc within the constrained nozzle is described. Furthermore, the powder feed rate, as a key variable determining coating thickness, and its interaction with the carrier gas pressure are incorporated into the model to establish the spatial distribution of sprayed particles in the jet and the temperature and velocity gradient laws upon impact with the workpiece surface.

[0035] After completing the basic modeling, a feature extraction algorithm is used to correlate the end-effector linear velocity, angular velocity, and attitude angle change rate of the robotic arm with the deposition efficiency and coating thickness distribution of plasma spraying. In this embodiment, the nonlinear mapping model adopts a deep residual network structure. The input layer of this network receives the normalized motion state vector of the robotic arm, including the current Cartesian coordinates, instantaneous velocity vector, and attitude offset relative to the workpiece normal, while also introducing environmental disturbance factors, such as the ambient humidity, ambient pressure, and instantaneous wind speed of the exhaust system in the spraying chamber. The hidden layer extracts the deep logical correlation between these motion features and the process results through multi-layer nonlinear transformations. The output layer provides suggested values ​​for process parameter correction for the current motion state. The introduction of the deep residual network effectively solves the gradient vanishing problem in the training process of deep networks, enabling the model to learn the microsecond-level evolution of the spraying energy distribution during the rapid acceleration and deceleration phases of the robotic arm.

[0036] Step S2 is executed to acquire the three-dimensional geometric information of the workpiece to be processed. The system utilizes a high-precision three-dimensional vision sensor integrated at the end of the robotic arm or fixed above the workstation to perform a comprehensive scan of the turbine blade to be coated. The vision sensor emits structured light or laser stripes onto the blade surface and acquires millions of point cloud data by receiving the distortion of the reflected light. After noise reduction processing, isolated noise points caused by reflections from the metal surface are removed from these raw data. Subsequently, a surface reconstruction algorithm is used to transform the discrete point cloud into a continuous non-uniform rational spline surface.

[0037] On the reconstructed surface, equally spaced scanning paths are planned based on preset coating thickness parameters (e.g., 120 micrometers) and overlap requirements (e.g., 50%). For high curvature areas such as the inlet and outlet edges of the blades, the system automatically densifies the path distribution and accurately calculates the normal vector information at each path point. Based on the curvature variation pattern of the path points, the target motion speed of the robotic arm and the rated power of the plasma spray gun are preset at each path point. In areas with high curvature, the linear velocity of the robotic arm is usually reduced to maintain smooth posture switching. At this time, the model automatically adjusts the arc power and powder feeding rate to prevent excessive local coating thickness. In the flat area in the middle of the blade, the robotic arm operates at a higher speed, and the process parameters are increased proportionally. Finally, this information is integrated into a complete process path guidance file, serving as the logical foundation for subsequent operations.

[0038] In step S3, real-time data is collected across all dimensions during the operation. Through the high-speed synchronous data interface provided by the robotic arm controller, the system acquires the absolute position feedback values ​​of each joint encoder at a sampling frequency of over 200 Hz. This data undergoes forward kinematic transformation to obtain the real-time values ​​of the end effector's three translational degrees of freedom (lateral, longitudinal, and vertical) and three rotational degrees of freedom (roll, pitch, and yaw) in the Cartesian coordinate system.

[0039] Meanwhile, real-time output parameters of the plasma power supply are acquired via industrial fieldbuses (such as industrial Ethernet buses), including instantaneous output voltage, current ripple, and inlet and outlet temperatures and flow rates of cooling water in the heat exchange system. These parameters directly reflect the stability of the plasma arc. The instantaneous rotational speed of the powder feeder and the pressure fluctuations of the powder-feeding carrier gas are also recorded in real time to monitor the uniformity of the powder feeding process. To ensure control synchronization, all acquired data are aligned on a unified high-precision time scale to ensure zero-drift coupling between motion data and process data in the time dimension.

[0040] Step S4 is executed, entering the dynamic calculation and adjustment stage. The system first calculates the spatial deviation between the current actual running trajectory and the preset path. When trajectory lag or overshoot caused by the dynamic characteristics of the robotic arm is detected, the inverse kinematics algorithm is used to solve the compensation torque that needs to be added to each joint in real time. When dynamically calculating the motion speed compensation of the robotic arm, a predictive control algorithm is introduced. This algorithm is not limited to the current error feedback, but predicts the trajectory trend for the next 5 to 10 sampling periods based on the current motion state vector. By minimizing the objective function (i.e., the sum of the deviations between the preset trajectory and the predicted trajectory) in the time domain, the optimal control increment is solved, thereby preemptively offsetting the lag caused by the mechanical inertia of the system.

[0041] On the process side, based on fluctuations in the instantaneous spraying distance (caused by robotic arm vibration or workpiece installation tolerances), a nonlinear mapping model is used to correct the plasma arc power in real time. If the spray gun gets closer to the workpiece surface, the system controls the power module to fine-tune the current output via a high-speed communication protocol, and simultaneously fine-tunes the ratio of main gas to auxiliary gas to change the compression degree of the plasma arc, thereby achieving dynamic control of the spray beam diameter and ensuring that the particle impact energy density remains constant. When the robotic arm decelerates at a curve in its trajectory, the system synchronously adjusts the stepper motor speed of the powder feeder according to the percentage decrease in real-time linear velocity, ensuring that the amount of powder deposited per unit length of path is not affected by changes in movement speed.

[0042] Step S5 is executed to achieve closed-loop control with multi-dimensional sensing feedback. The system uses an infrared thermal imaging sensor to monitor the instantaneous temperature field on the blade surface in real time. When the infrared thermal imager detects that the temperature in a local area exceeds the preset ceramic powder sintering threshold or the heat resistance limit of the substrate due to heat accumulation, the central integrated control module will immediately make a decision: either trigger the robotic arm to temporarily increase its operating speed without changing its trajectory, or directly reduce the arc energy output of the plasma power supply.

[0043] Simultaneously, the system outputs a prediction of the coating growth state. This prediction is based on a numerical simulation algorithm, combined with currently collected data on current, voltage, airflow, powder flow, and real-time spraying distance and angle. It can accumulate and calculate the deposition amount of each coating layer in real time and establish a digital twin model to simulate the physical stacking process of the coating. Before the entire workflow is completed, the system can provide a preliminary assessment of the coating's density and a prediction of its bonding strength with the substrate, providing a basis for quality self-diagnosis.

[0044] Example 2 This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 Specifically, step S1 involves establishing a nonlinear mapping model between the motion characteristics of the multi-axis robotic arm and the plasma spraying process parameters, including: A dynamic model considering the inter-joint friction torque, centripetal torque, and Coriolis torque of a multi-axis robotic arm is established using the Lagrange equation to predict the inertial response characteristics and trajectory tracking error of the robotic arm when performing highly dynamic actions. A physical characteristic model of the plasma spray gun is established to describe the influence of the fluctuation of arc voltage and arc current on the degree of plasma ionization. Combined with the main gas flow rate, auxiliary gas flow rate and powder feeding rate, the spatial distribution law of sprayed particles in the plasma jet and the energy gradient law when impacting the workpiece surface are established. Using feature extraction algorithms, the end-effector linear velocity, angular velocity, and attitude angle change rate of the robotic arm are used as inputs to a deep residual network structure. Environmental disturbance factors are introduced into the hidden layer for nonlinear feature correlation extraction, and the output layer provides suggested values ​​for process parameter correction for the current motion state.

[0045] Step S2, which involves obtaining the three-dimensional geometric information of the workpiece to be processed and generating a set of path points, includes: A 3D vision sensor is used to emit structured light or laser stripes onto the surface of the workpiece. The original point cloud data is obtained by receiving the distortion of the reflected light, and noise reduction processing is performed on the original point cloud data to remove isolated noise points. Discrete point cloud data is transformed into continuous non-uniform rational spline surfaces through surface reconstruction algorithms. Equally spaced scanning paths are planned on the reconstructed surfaces, and the path distribution is densified for high curvature regions to accurately calculate the normal vector information at path points. Based on the curvature changes of the path points, the attitude switching logic of the robotic arm at different positions is determined, and the power output of the plasma spray gun is synchronously preset according to the preset trend of the robotic arm's linear velocity, generating a process path guidance file containing spatial coordinates, attitude vectors and process thresholds.

[0046] In this embodiment, by establishing a nonlinear mapping relationship between the motion characteristics of the multi-axis robotic arm and the plasma spraying process parameters in step S1, the trajectory deviation and end-effector posture fluctuation generated by the robotic arm during high-speed operation, variable speed operation, and posture switching are correlated with the arc power, airflow parameters, and powder feeding parameters of the plasma spray gun. This allows the changes in the robotic arm's motion state and the adjustment of the spraying process to be processed under the same control link. In step S2, by acquiring the three-dimensional geometric information of the workpiece to be processed and performing surface reconstruction, path point planning, normal vector extraction, and high curvature area path densification processing, a process path guidance file containing spatial coordinates, posture vectors, and corresponding process setting information is generated. This enables the robotic arm to call the corresponding posture switching mode and power output mode according to the curvature changes in different areas when performing spraying along the workpiece surface. In the above manner, when facing irregular curved surfaces, edge transition areas, and areas with abrupt curvature changes, there is a continuous correspondence between the robotic arm's motion trajectory and the spraying process parameters. Problems such as local deposition imbalance at the spraying path switching position, discontinuous boundary transition, and asynchronous attitude adjustment and process output can be addressed in a targeted manner, making it easier for subsequent steps to continue to perform real-time acquisition, dynamic correction, and closed-loop control according to the path point information.

[0047] Example 3 This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 Specifically, step S3 involves real-time acquisition of the robotic arm's pose data and the instantaneous operating parameters of the plasma spray gun, including: The high-speed data interface of the robotic arm controller reads the encoder feedback values ​​of each joint and converts the joint position data into the pose data of the end effector in the Cartesian coordinate system based on the positive kinematics of the robotic arm. The pose data includes at least the position of the end effector in the X, Y and Z directions, as well as the attitude angle around the corresponding coordinate axis. To ensure that the motion data and subsequent process data are under the same control cycle, the data reading cycle is kept consistent with the robotic arm interpolation cycle. The output voltage, output current, main gas flow rate, auxiliary gas flow rate, powder feeder speed, powder carrier gas pressure, and cooling water inlet and outlet temperatures of the plasma power supply are synchronously read via industrial fieldbus. The read process data is written into the instantaneous operating status record sequence according to the same time base as the pose data. Among them, the output voltage and output current are used to characterize the current energy state of the plasma arc, the main gas flow rate and auxiliary gas flow rate are used to characterize the jet compression state, and the powder feeder speed and powder carrier gas pressure are used to characterize the powder conveying state. The spraying distance offset and spraying angle deviation are obtained by a detection component set at the end of the robotic arm or near the spray gun. The robotic arm pose data, plasma spray gun operating status parameters, spraying distance offset and spraying angle deviation are uniformly timestamped to form a synchronous monitoring data group under a unified time scale. This allows subsequent steps to perform joint analysis based on the robotic arm motion state and spraying process state at the same time.

[0048] The process of dynamically adjusting the robotic arm's motion speed compensation and the plasma spraying system's execution variables in step S4 includes: The synchronous monitoring data set is compared with the process path guidance file generated in step S2 on a time-by-time basis. The target spatial coordinates, target attitude vector and target process parameters of the corresponding path point at the current time are read, and the spatial deviation between the actual pose of the robotic arm end and the target pose is calculated. The spatial deviation includes position deviation and attitude deviation. Based on the spatial deviation, the compensation amount of each joint is solved by inverse kinematics, and the compensation amount is written into the motion control channel of the robotic arm to correct the actual running trajectory. On this basis, the trajectory evolution trend in the next few sampling periods is predicted according to the current motion state, and the corresponding motion control increment is given in advance to deal with the lag response generated by the robotic arm in the acceleration, deceleration and attitude switching phases. While compensating for the movement of the robotic arm, based on the current spraying distance offset, spraying angle deviation, and current linear velocity change, the nonlinear mapping model established in step S1 is invoked to output process parameter correction suggestions corresponding to the current motion state. These process parameter correction suggestions are then written into the plasma power control channel, gas path control channel, and powder feeding control channel to synchronously correct the arc power, main gas flow rate, auxiliary gas flow rate, and powder feeding rate, ensuring that the spray gun execution state corresponds to the robotic arm end-effector motion state.

[0049] In this embodiment, by synchronously collecting the pose data of the robotic arm and the instantaneous operating parameters of the plasma spray gun, and performing pose deviation analysis and process parameter linkage correction under a unified time reference, the robotic arm movement side and the spraying process side can be processed in the same control moment, which facilitates continuous correction of trajectory deviation, spraying distance fluctuation and powder feeding cycle change in the subsequent spraying process.

[0050] Example 4 This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 and Figure 3 Specifically, step S5, which involves multi-dimensional sensing feedback to achieve closed-loop control and output prediction results, includes: The real-time temperature image of the workpiece surface is acquired using an infrared thermal imaging sensor, and the real-time temperature image is divided into regions to form surface temperature distribution data corresponding to the path point set in step S2. The instantaneous temperature value, local temperature expansion direction, and continuous temperature rise status of each region are read from the temperature distribution data to identify whether there are heat concentration sections in the current spraying process. When a localized heat concentration area is identified, the central integrated control module combines the location of the corresponding path point, the real-time linear speed of the robotic arm, the power output status of the spray gun, and the spraying distance offset status to select whether to adjust the robotic arm's running speed, adjust the plasma power supply's output power, or switch the order of subsequent paths to be executed. Specifically, when the heat concentration area is located within the current continuous path, the robotic arm's running speed or spray gun power is adjusted first; when the heat concentration area is located in the path intersection area, the execution order of subsequent paths is adjusted first. A digital twin model is established to run synchronously with the actual spraying process. The output voltage, output current, main gas flow rate, auxiliary gas flow rate, powder feeding rate, spraying distance offset, spraying angle deviation, and real-time temperature data of the workpiece surface collected in step S3 are continuously written into the digital twin model. The digital twin model is used to track and calculate the particle deposition process to obtain the current coating thickness accumulation state, local deposition continuity state, and coating growth state of the entire spraying area. Based on the cumulative state of coating thickness, the continuous state of local deposition, and the temperature distribution state, a comprehensive judgment is made on the completed spraying area and the area to be sprayed, and a prediction result of the coating growth state is output. The prediction result includes at least the thickness development trend of the current area, the risk state of local heat accumulation, and the process call suggestions for subsequent path segments. The prediction result is fed back to the central integrated control module to correct the process execution strategy corresponding to the subsequent path points.

[0051] Furthermore, in this embodiment, the method may also include the following supplementary control procedures: After generating the set of robotic arm operation path points, for workpiece edge areas and curvature change areas, the trend of the angle change between the spray gun axis and the workpiece surface normal is first identified. Then, the posture transition mode between adjacent path points is smoothed according to the angle change trend, so that the posture change process of the robotic arm at the path transition position remains continuous. After the posture transition is completed, the nonlinear mapping model is called to recalculate the process parameters corresponding to the transition segment, so that the spraying state of the posture transition segment is connected with the adjacent normal path segment. For cases where the powder feeding system responds later than the control command changes, the time difference between the time when the powder feeding control command is issued and the time when the powder actually arrives at the spray gun outlet is recorded in advance. Before the robotic arm is about to enter the acceleration section, deceleration section or curvature change area, the powder feeding adjustment command is issued in advance according to the time difference, so that the time of the powder feeding state change corresponds to the position of the change in the linear velocity of the robotic arm. When processing environmental disturbances in layers, the power supply side is first corrected for arc state fluctuations, the motion side is then corrected for pose changes caused by workpiece installation deviations or thermal deformation, and finally the execution sequence of the spraying path is adjusted according to the overall thermal distribution state of the workpiece surface, so that the spraying process forms corresponding adjustment relationships in the process layer, motion layer and path layer respectively.

[0052] In this embodiment, by introducing workpiece surface temperature field monitoring, digital twin deposition calculation, powder advance adjustment, and path sequence switching processing, the spraying process can maintain continuous control when thermal state changes, path switching, and material delivery cycle changes, which facilitates the linkage processing of the deposition state of local areas and the execution state of subsequent paths.

[0053] Example 5 An integrated control system for a robotic arm plasma spray gun, please refer to... Figure 2 Specifically: The system includes a central integrated control module, a multi-axis robotic arm drive module, a plasma process execution module, and a sensing module. The central integrated control module is used to execute the nonlinear mapping model construction in step S1, the process path guidance file generation in step S2, the synchronous data processing in step S3, the motion compensation and process variable linkage correction in step S4, and the closed-loop interpretation and coating growth state prediction in step S5. The central integrated control module internally stores the process path guidance file, the instantaneous running status record sequence, the synchronous monitoring data group, and the coating growth state prediction results, and controls each module to perform data interaction according to a predetermined time sequence through a unified scheduling logic. The multi-axis robotic arm drive module is connected to the central integrated control module and is used to receive path execution instructions, attitude switching instructions and compensation control instructions output by the central integrated control module, and drive the spray gun at the end of the robotic arm to perform spraying motion according to the spatial coordinates and attitude vectors corresponding to the process path guidance document; wherein, the multi-axis robotic arm drive module is also used to transmit the feedback values ​​of each joint encoder back to the central integrated control module in real time. The plasma process execution module is connected to the central integrated control module and is used to receive arc power adjustment commands, main gas flow rate adjustment commands, auxiliary gas flow rate adjustment commands, and powder feeding rate adjustment commands output by the central integrated control module. It also controls the execution status of the plasma power supply, gas path control cabinet, and powder feeder according to the adjustment commands. The plasma process execution module is also used to transmit output voltage, output current, airflow status, carrier gas pressure, powder feeder speed, and cooling status parameters back to the central integrated control module. The sensing module is connected to the central integrated control module and is used to acquire the three-dimensional point cloud data of the workpiece to be processed, the spraying distance offset, the spraying angle deviation, and the real-time temperature image of the workpiece surface during the spraying process, and send the acquired data to the central integrated control module; wherein, the sensing module includes at least a three-dimensional vision sensor, a pose detection component, and an infrared thermal imaging sensor.

[0054] In this embodiment, the central integrated control module first calls the sensing module to complete the point cloud acquisition and surface reconstruction of the workpiece surface, and generates a process path guidance file; then it calls the multi-axis robotic arm drive module and the plasma process execution module to enter the actual spraying state; during the spraying process, the central integrated control module continuously receives data from the multi-axis robotic arm drive module, the plasma process execution module and the sensing module, and calculates the robotic arm compensation amount and process variable correction amount based on the synchronous monitoring data group; when it is detected that the workpiece surface temperature distribution is abnormal or the coating growth state deviates from the predetermined state, the central integrated control module continues to output new control commands to the multi-axis robotic arm drive module and the plasma process execution module to adjust the execution state of the current spraying process or subsequent path segments.

[0055] In this embodiment, the system uses a central integrated control module as the data aggregation center and control output center, so that the robotic arm motion control link, plasma process control link and sensor feedback link form a unified data interaction relationship, which facilitates the execution path calling, state correction and closed-loop control according to the same time base.

[0056] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for integrated control of a robotic arm plasma spray gun, characterized in that: Includes the following steps: S1. Based on the dynamic characteristics of multi-axis robotic arms and the physical evolution law of plasma spraying, a nonlinear mapping model of the motion characteristics of multi-axis robotic arms and plasma spraying process parameters is constructed. Among them, the end-effector jitter characteristics and trajectory deviation distribution of the robotic arm under different load and acceleration conditions are obtained by establishing a dynamic state space expression. At the same time, the influence relationship of process parameters on the velocity and temperature of sprayed particles is determined by analyzing the arc energy conversion and jet dynamics inside the plasma spray gun. The end-effector motion state vector of the robotic arm is correlated with the deposition efficiency of plasma spraying using a deep residual network structure. S2. Use a 3D vision perception mechanism to acquire surface point cloud data of the workpiece to be processed and perform surface reconstruction. According to the preset coating thickness index and path overlap requirements, plan a set of path points on the reconstructed surface and extract the normal vector information of each path point. Based on the curvature change law of the path points, preset the target motion speed of the robotic arm and the rated power of the plasma spray gun for each path point to form a process path guidance document. S3. During the operation, the encoder feedback values ​​of each joint of the robotic arm are collected in real time through the high-speed data interface and converted into the position and pose data of the end effector in the Cartesian coordinate system. The instantaneous operating status parameters of the plasma spray gun, including the output parameters of the plasma power supply, the air path status and the powder feeding status, are simultaneously acquired, and all collected data are aligned on a unified time scale line. S4. Based on the mapping relationship between motion characteristics and process parameters, calculate the spatial deviation between the actual running trajectory and the preset path, solve the compensation torque of each joint of the robotic arm through the inverse kinematics algorithm, and synchronously correct the execution variables of the plasma spraying system according to the fluctuation of instantaneous spraying distance and the change of the linear velocity of the robotic arm. S5. Monitor the instantaneous temperature field distribution on the workpiece surface through a multi-dimensional sensing feedback mechanism, execute closed-loop thermal balance control of the spraying process, and calculate the coating deposition amount in real time based on numerical simulation algorithm, and output the prediction results of the coating growth state.

2. The integrated control method for a robotic arm plasma spray gun according to claim 1, characterized in that: The process of establishing a nonlinear mapping model between the motion characteristics of the multi-axis robotic arm and the plasma spraying process parameters in step S1 includes: A dynamic model considering the inter-joint friction torque, centripetal torque, and Coriolis torque of a multi-axis robotic arm is established using the Lagrange equation to predict the inertial response characteristics and trajectory tracking error of the robotic arm when performing highly dynamic actions. A physical characteristic model of the plasma spray gun is established to describe the influence of the fluctuation of arc voltage and arc current on the degree of plasma ionization. Combined with the main gas flow rate, auxiliary gas flow rate and powder feeding rate, the spatial distribution law of sprayed particles in the plasma jet and the energy gradient law when impacting the workpiece surface are established. Using feature extraction algorithms, the end-effector linear velocity, angular velocity, and attitude angle change rate of the robotic arm are used as inputs to a deep residual network structure. Environmental disturbance factors are introduced into the hidden layer for nonlinear feature correlation extraction, and the output layer provides suggested values ​​for process parameter correction for the current motion state.

3. The integrated control method for a robotic arm plasma spray gun according to claim 1, characterized in that: Step S2, which involves obtaining the three-dimensional geometric information of the workpiece to be processed and generating a set of path points, includes: A 3D vision sensor is used to emit structured light or laser stripes onto the surface of the workpiece. The original point cloud data is obtained by receiving the distortion of the reflected light, and noise reduction processing is performed on the original point cloud data to remove isolated noise points. Discrete point cloud data is transformed into continuous non-uniform rational spline surfaces through surface reconstruction algorithms. Equally spaced scanning paths are planned on the reconstructed surfaces, and the path distribution is densified for high curvature regions to accurately calculate the normal vector information at path points. Based on the curvature changes of the path points, the attitude switching logic of the robotic arm at different positions is determined, and the power output of the plasma spray gun is synchronously preset according to the preset trend of the robotic arm's linear velocity, generating a process path guidance file containing spatial coordinates, attitude vectors and process thresholds.

4. The integrated control method for a robotic arm plasma spray gun according to claim 1, characterized in that: Step S3, which involves real-time acquisition of robotic arm pose data and instantaneous operating parameters of the plasma spray gun, includes: By utilizing the high-speed synchronous data interface of the robotic arm controller, the absolute position feedback values ​​of the encoders of each joint are obtained. The translational and rotational degrees of freedom of the end effector in the Cartesian coordinate system are obtained through forward kinematic transformation, while ensuring that the data update frequency is synchronized with the interpolation cycle of the robotic arm. The instantaneous output voltage and current ripple of the plasma power supply, as well as the inlet and outlet temperatures of the cooling water in the heat exchange system, are acquired in real time through the industrial fieldbus. The instantaneous speed of the powder feeder and the pressure fluctuation of the powder carrier gas are also recorded. Sensors installed at the end of the robotic arm monitor the spraying distance offset and spraying angle deviation. A central clock source timestamps all sensor signals to eliminate the coupling offset between motion data and process data in the time dimension.

5. The integrated control method for a robotic arm plasma spray gun according to claim 1, characterized in that: The process of dynamically adjusting the robotic arm's motion speed compensation and the plasma spraying system's execution variables in step S4 includes: The real-time pose data is compared with the process path guidance file to calculate the spatial pose error of the actual trajectory relative to the target trajectory. The inverse kinematics algorithm is then used to convert the spatial pose error into compensation current commands for each joint. Predictive control algorithms are introduced during dynamic adjustment. Based on the current motion state vector, the trajectory evolution trend of multiple future sampling periods is predicted. By minimizing the sum of the deviations between the target trajectory and the predicted trajectory in the time domain, the optimal control increment is solved to offset the hysteresis caused by mechanical inertia. Based on the real-time monitoring of the spraying distance fluctuations, the arc power is corrected through a nonlinear mapping model, and the flow ratio of the main gas and auxiliary gas is simultaneously fine-tuned to adjust the compression degree of the plasma arc, thereby controlling the physical envelope range of the spraying beam.

6. The integrated control method for a robotic arm plasma spray gun according to claim 1, characterized in that: Step S5, the process of achieving closed-loop control and outputting predicted results through multi-dimensional sensing feedback, includes: The infrared thermal imaging sensor is used to collect real-time temperature images of the workpiece surface. The image processing algorithm identifies the local heat accumulation. When the local temperature exceeds the preset thermal damage threshold, the robot arm’s running speed is automatically adjusted or the energy output of the plasma power supply is reduced. Establish a digital twin model synchronized with the physical operation process, and combine real-time collected current, voltage, airflow and powder flow data to simulate the particle stacking process on the workpiece surface and calculate the real-time cumulative amount of the current coating thickness. Based on the deviation between the coating physical stacking model and real-time process parameters, the coating density level and bonding strength after the operation are predicted, and the prediction results are fed back to the central integrated control module to correct the spraying execution strategy of subsequent paths.

7. The integrated control method for a robotic arm plasma spray gun according to claim 1, characterized in that: When generating the set of work path points for the robotic arm, the following path optimization logic is executed for the edge regions and geometrically abrupt locations of the workpiece: Collision detection algorithms are used to identify potential interference risks between the robotic arm body and the workpiece edge. Under the premise of ensuring that the verticality deviation of the spray gun is within the allowable range, the joint configuration is finely adjusted by utilizing the redundant degrees of freedom of the robotic arm. The trajectory inflection points are smoothed by using spline curve interpolation, and sharp angle switching is transformed into a circular arc transition trajectory to suppress the impact vibration of the robotic arm during the reversal process. The process compensation amount under the new trajectory is recalculated by nonlinear mapping model to ensure that the amount of material deposited per unit length of the path remains constant during the smooth transition phase of the path.

8. The integrated control method for a robotic arm plasma spray gun according to claim 1, characterized in that: The method also involves advance compensation control logic for powder delivery lag: Based on the physical length of the powder feeding pipeline and the carrier gas velocity, the physical delay time of the powder feeding system from command change to outlet flow response is pre-calibrated; The central integrated control module predicts the acceleration and deceleration points of the robotic arm on the trajectory and, in conjunction with the physical delay time, outputs a powder feeding flow adjustment command at a specific moment before the robotic arm reaches the predetermined motion state change point. The time synchronization mechanism ensures that the fluctuation points of powder flow rate and the changes in the linear velocity of the robotic arm coincide in spatial position.

9. The integrated control method for a robotic arm plasma spray gun according to claim 1, characterized in that: During closed-loop control, the system implements a multi-level environmental disturbance suppression mechanism: The first level provides millisecond-level power supply voltage and current regulation compensation for arc fluctuations; The second level uses visual monitoring data to correct pose deviations caused by workpiece thermal deformation or installation tolerances in real time, and compensates them in the robot arm motion coordinate system. The third level monitors the overall thermal balance of the workpiece surface and dynamically schedules the execution sequence of the spraying path. When a local high temperature is detected, the robotic arm is guided to jump to a low-temperature area to perform the operation, thus realizing thermal intervention at the path level.

10. An integrated control system for a robotic arm plasma spray gun, applied to the integrated control method for a robotic arm plasma spray gun as described in any one of claims 1 to 9, characterized in that: The system includes: The central integrated control module, with its built-in high-speed logic processor and large-capacity real-time storage unit, is used to perform nonlinear mapping model calculations, path planning, and multi-protocol data parsing, realizing the deep integration of robotic arm control logic and plasma process control logic. The multi-axis robotic arm drive module has its input end connected to the motion command output end of the central integrated control module, and has a hardware-level synchronization enable interface for receiving motion compensation commands and driving the robotic arm to execute a predetermined trajectory, ensuring high synchronization between the start of spraying and the start of motion. The plasma process execution module includes a high-frequency inverter plasma power supply, a gas path control cabinet, and a closed-loop powder feeder. The components communicate with each other through a real-time deterministic network and are controlled by the process adjustment signals output by the central integrated control module to perform dynamic fine-tuning of jet energy and material flow rate. The sensing module includes a high-precision encoder, a 3D vision sensor, an infrared thermal imaging sensor, and a pressure sensor. It is used to monitor the system's operating status in all dimensions in real time and transmit the signals back to the central integrated control module. The feedback data from different dimensions is weighted and noise filtered through a data fusion algorithm.

Images