Continuous feeding manipulator control method and system

Abstract

The invention discloses a continuous feeding manipulator control method which comprises the following steps: constructing a finite element simulation model aiming at each joint of a mechanical arm and a tail end execution structure, acquiring inherent frequency distribution data of the structure under different load states and posture working conditions, and generating a frequency sensitive area parameter set based on the data. The method has the advantages that the track segment exciting the resonance risk can be recognized before the path is executed, the inherent frequency of the structure is effectively avoided through frequency matching and path reconstruction, the dynamic adaptability is improved, and pose offset and structural fatigue caused by resonance are prevented. Meanwhile, the system has the self-adaptive closed-loop optimization capacity, disturbance injection testing and periodic response verification are combined, the structure state can be monitored in real time, path re-optimization can be triggered, the stability and precision of the mechanical arm in continuous high-frequency operation are ensured, and the overall reliability and intelligent level of a production line are improved.

Description

Technical Field

[0001] The present invention relates to the field of automation control technology, and in particular to a continuous feeding robot control method and system. Background Art

[0002] "Continuous loading robot control" refers to the operation method of controlling the robot to transport raw materials or workpieces from the feed area to the processing equipment or assembly position in a continuous and uninterrupted manner during the automated production process. Compared with traditional intermittent loading, this control method emphasizes that in scenarios where the material input rhythm is fast or the production efficiency requirements are high, by optimizing the robot's motion path, grasping rhythm, material identification and caching mechanism, it can achieve rapid connection and seamless transmission of materials and avoid pauses caused by waiting, misalignment or interference. This method usually integrates technologies such as sensor recognition, multi-axis coordinated control, and dynamic path planning to improve loading efficiency and the overall stability of the production line. It is suitable for scenarios with high requirements for rhythm continuity, such as automated production lines, injection molding loading and unloading, and CNC machining.

[0003] In the existing control process of continuous-feeding robots, with the continuous increase in operating rhythm and refinement of path instructions, there is a risk that some segments of the robot's trajectory, when operating continuously at high speed, may excite the natural frequency of the structure. Specifically, if the trajectory parameters (such as path curvature, acceleration pulse frequency, and repetition rhythm) are close to the natural frequency of the robot body or end-effector structure, a low-amplitude, high-frequency resonant response may be induced in the absence of external interference. Although the initial amplitude of this resonance is small, it has a cumulative characteristic during continuous operation, which can easily lead to a decrease in the accuracy of end-position control, micro-deviations in the motion path, and even cause structural loosening, joint component fatigue, and servo system misidentification. In more serious cases, the resonant disturbance will be amplified through the mechanical structure layer by layer, interfering with the synchronous scheduling of other coordinated mechanisms, causing rhythm disruptions, misgrabbing, collisions, or sudden stops on the entire production line, seriously affecting the stability and reliability of the system. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a continuous feeding robot control method and system.

[0005] In order to solve the above technical problems, the technical solution of the present invention is: a continuous feeding robot control method, comprising the following steps: Construct finite element simulation models for each joint and end-effector structure of the robotic arm, obtain the natural frequency distribution data of the structure under different load states and posture conditions, and generate a set of frequency-sensitive region parameters based on the data; The robot's planned running path is segmented and expanded into a time-domain excitation signal sequence. The excitation frequency characteristics of each trajectory segment in the sequence are calculated. The frequency characteristics are matched with the frequency-sensitive region parameter set to identify the trajectory segments that have the risk of exciting structural resonance. For the identified trajectory segments with resonance excitation risks, a time-domain disturbance response model is constructed to simulate the disturbance propagation path and response amplitude of the current trajectory segment during continuous operation. The offset of the trajectory disturbance on the terminal posture accuracy is calculated, and a resonance cumulative risk index is generated based on the response trend. Based on the resonance cumulative risk index, dynamic trajectory path reconstruction is performed, and the acceleration pulse sequence of the target trajectory is adjusted using a nonlinear adjustment function to generate a low-frequency coupling path sequence that avoids the natural frequency of the structure. The low-frequency coupling path sequence is input into the manipulator scheduling module. A dynamic trajectory verification mechanism is set up in the early stage of control system operation. The structural response is tested by micro-injection of trajectory disturbances, and the path control parameters are corrected in real time based on the response data. During the operation of the control system, periodic structural response verification operations are performed to collect data on the end-point posture accuracy fluctuation, path execution residuals, and mechanical joint drive current changes to evaluate the stability of the operating state. If it is determined that there is a response abnormality, the path re-optimization operation is triggered.

[0006] Preferably, the specific steps of constructing a finite element simulation model for each joint and end effector structure of the robotic arm are as follows: A finite element model is built based on the structural components of the robotic arm. The model includes the base, multiple joints, link units, and end fixtures, and meshes are divided according to size, material properties, and connection methods. Set up multiple groups of posture and load conditions, apply boundary conditions and loads to each group of conditions, perform modal analysis, and extract modal frequencies and vibration mode distributions; The frequency band clustering method is used to statistically process the modal data, extract the active response sections, and construct the frequency sensitive area parameter set; The frequency-sensitive area parameter set is stored in the frequency feature database module, and trajectory resonance risk identification and dynamic update are supported.

[0007] Preferably, the specific steps of segmenting the predetermined running path of the manipulator into a time domain excitation signal sequence and identifying the trajectory segment that induces structural resonance are as follows: Obtain the robot's loading path data once, and construct a time-domain excitation signal sequence containing acceleration, velocity, and displacement changes in segments according to the time dimension; Perform frequency domain analysis on each trajectory segment to extract the excitation frequency characteristics, including the main frequency, secondary frequency, frequency energy distribution and frequency change trend; Match the frequency features with the frequency-sensitive region parameter set to identify trajectory segments with overlapping frequencies and energy proportions exceeding a threshold as those with structural excitation risks. Data archiving and visualization of the identified trajectory segments are performed, and a risk identification matrix is ​​generated for path reconstruction.

[0008] Preferably, the specific steps of constructing a time-domain disturbance response model for the identified trajectory segment with the risk of resonance excitation are as follows: Extract the dynamic control parameters of the resonance risk trajectory segment, construct the acceleration excitation function and current pulse as the disturbance input, establish the multi-body dynamic model and perform time domain response calculation; Perform time-stepping simulation to output terminal pose drift, structural strain and torque changes, analyze disturbance propagation paths, and construct a mapping relationship between path parameters and structural responses; Based on the simulation results, the trajectory disturbance risk factor, joint stress risk value, and servo response deviation are calculated and standardized into a unified risk index. Integrate risk indicators into the trajectory scheduling system to generate path risk profiles and mark priority segments for path intervention.

[0009] Preferably, the specific steps of performing dynamic trajectory path reconstruction based on the resonance cumulative risk index are as follows: The trajectory segments are sorted according to the trajectory disturbance risk factor and the cumulative coefficient of joint stress amplitude, and the high-risk trajectory segments and their original path parameters are extracted as reconstruction targets; A nonlinear regulation function is constructed to adjust the acceleration pulse sequence, and the excitation main frequency is shifted to outside the natural frequency of the structure through frequency amplitude transfer. Perform frequency domain analysis and calculate the frequency coupling suppression coefficient. If the energy ratio is lower than the threshold, the reconstruction is considered to meet the standard. If it does not meet the standard, re-optimize; The reconstructed path is embedded in the scheduling module and the servo execution response data is recorded, and the path version database is updated.

[0010] Preferably, the steps of inputting the low-frequency coupling path sequence into the manipulator scheduling module and setting a dynamic trajectory verification mechanism include: Import the low-frequency coupling path sequence into the trajectory scheduling module, complete timestamp mapping and buffer queue management, and load it into the task block execution logic; During the initialization phase, a small-amplitude disturbance signal is injected into multiple path nodes and the terminal position drift, acceleration change, and current fluctuation response data are collected; Triggering the path parameter correction mechanism based on the response deviation to perform acceleration slope adjustment, inter-node time delay, or servo response sensitivity reduction operations; The verified path mark is archived, and after simulation verification, it is determined whether to write it into the path version database as a structural correction adaptation path.

[0011] Preferably, in order to ensure the structural response stability of the manipulator during long-term operation in the continuous loading process, a periodic structural response verification mechanism is proposed to regularly collect key response data during the control system operation cycle, comprehensively evaluate the system operation status, and automatically trigger the path re-optimization operation when an abnormality is detected. The specific steps are as follows: During the operation of the control system, the periodic acquisition of structural response data is performed in units of a fixed working cycle. The key parameters collected include the end-point pose accuracy fluctuation, the path execution residual, and the rate of change of the mechanical joint drive current. The structural stability monitoring factor is calculated based on the acquired parameters. The calculation expression is as follows: , where It is the average offset between the actual position and theoretical position of the end of the robot during the execution of the path. is the cumulative error between the actual execution trajectory and the target trajectory, is the ratio of the current fluctuation in the previous cycle to that in the previous cycle, 、 as well as They are 、 as well as The weight coefficient of is the structural stability monitoring factor; Obtaining the structural stability monitoring factor of the current cycle Then, combine it with the previous Compare the mean values ​​within a period to determine whether there is an abnormal operating status. The calculation formula is: , where It is before Structural stability monitoring factor of the cycle, is the stability offset; If the calculated If it is greater than the set trigger threshold, it is determined that the current control system state has abnormal fluctuations, and the re-optimization trigger flag set by the control system will be set to 1: , where is the path reoptimization trigger threshold, It is the flag bit that triggers path re-optimization; When the path reoptimization trigger flag When activated, the control system will execute the path parameter disturbance correction process to generate a corrected new trajectory acceleration curve, as shown in the following formula: , where The target acceleration pulse of the original path, is the disturbance control gain coefficient; After the new track is generated, save it as a version , with the original path version For comparative testing, the new path will be collected again after actual operation 、 as well as value, calculate the new , compared with the old version, if the new stability index calculated after the path execution meets the following conditions: , the new path is considered to significantly improve the system response state, replaces the old path and enters production use; if the condition is not met, roll back to the old path and record the deviation data for subsequent optimization iterations, where is the structural stability monitoring factor, It is the structural stability monitoring factor under the original path.

[0012] A continuous loading manipulator control system includes a structural modal analysis module, a trajectory excitation feature recognition module, a disturbance response modeling module, a path reconstruction and optimization module, a trajectory dynamic verification module, and a structural stability monitoring and re-optimization module: The structural modal analysis module builds finite element simulation models for each joint and end-effector structure of the robotic arm, obtains the natural frequency distribution data of the structure under different load states and posture conditions, and generates a set of frequency-sensitive region parameters based on the data; The trajectory excitation feature recognition module breaks down the manipulator's planned path into segments and converts them into a time-domain excitation signal sequence. It then calculates the excitation frequency characteristics of each trajectory segment in the sequence, matches these frequency characteristics with a set of frequency-sensitive region parameters, and identifies trajectory segments that pose a risk of exciting structural resonance. The disturbance response modeling module constructs a time-domain disturbance response model for identified trajectory segments with resonance excitation risks. It simulates the disturbance propagation path and response amplitude of the current trajectory segment during continuous operation, calculates the offset of the trajectory disturbance on the terminal posture accuracy, and generates a resonance cumulative risk index based on the response trend. The path reconstruction optimization module performs dynamic trajectory path reconstruction based on the resonance cumulative risk index and uses a nonlinear adjustment function to adjust the acceleration pulse sequence of the target trajectory to generate a low-frequency coupling path sequence that avoids the natural frequency of the structure. The trajectory dynamic verification module inputs the low-frequency coupling path sequence into the manipulator scheduling module, sets up a dynamic trajectory verification mechanism in the early stage of control system operation, tests the structural response by micro-injecting trajectory disturbances, and performs real-time correction of path control parameters based on the response data; The structural stability monitoring and re-optimization module performs periodic structural response verification operations during the operation of the control system, collects data on end-position accuracy fluctuations, path execution residuals, and mechanical joint drive current changes, evaluates the stability of the operating state, and triggers path re-optimization operations if it is determined that there is a response abnormality.

[0013] Beneficial effects of the present invention: By constructing a finite element simulation model of the robotic arm structure and combining segmented analysis of the path excitation frequency with frequency-sensitive region matching, this invention can proactively identify trajectory segments with a risk of resonance excitation before the manipulator actually executes the path, thereby preventing the coupling of path excitation with the structural natural frequency. Compared to traditional control methods that rely on empirical settings or post-correction, this invention achieves feedforward identification and early intervention of dynamic structural responses, significantly improving the trajectory control system's dynamic adaptability and ability to prevent structural excitation issues. It effectively reduces system risks caused by resonance, such as structural fatigue, end-point precision drift, and joint control errors.

[0014] The present invention forms an adaptive closed-loop path optimization system through dynamic trajectory path reconstruction, disturbance micro-injection testing and periodic response verification mechanism. When the robot arm has trajectory execution deviation caused by structural aging, load changes or environmental disturbances in actual operation, the system can collect data such as end posture, path residual, current fluctuation in real time, and automatically trigger path re-optimization operation based on the structural stability monitoring factor. This dynamic path adjustment mechanism that integrates structural response feedback not only improves the stability and robustness of the control system in long-term continuous operation, but also enables the entire production line to have higher self-recovery capabilities and intelligent operation and maintenance level. It is particularly suitable for automated loading scenarios with high beats, long cycles and high precision positioning requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A method flow chart of a continuous feeding robot control method; Figure 2 A schematic diagram of the modules of a continuous loading robot control system. DETAILED DESCRIPTION

[0016] The following is a further description of specific embodiments of the present invention in conjunction with the accompanying drawings. It should be noted that the description of these embodiments is intended to facilitate understanding of the present invention and does not constitute a limitation of the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0017] Working principle of the present invention: A continuous feeding manipulator control method comprises the following steps: Construct finite element simulation models for each joint and end-effector structure of the robotic arm, obtain the natural frequency distribution data of the structure under different load states and posture conditions, and generate a set of frequency-sensitive region parameters based on the data; A structural frequency model is constructed to analyze the resonance risk that may occur in the robot arm during continuous loading control. Finite element analysis is used to model the robot arm's joint units, connecting rod structures, and end-effector structures. Multi-scenario simulations are then performed on their natural frequency distributions based on actual operating conditions to generate a set of frequency-sensitive region parameters that can be used for path risk identification. This implementation includes the following four steps: First, a global finite element structural model is established based on the structural components of the selected industrial robot. This includes the robot's base, multiple rotational or linear joints, linkage units, and fixtures, suction cups, or specialized tooling associated with the end-effector's task. During the modeling process, the solid geometry is constructed based on factors such as the actual robot's dimensions, material properties (such as density, Young's modulus, Poisson's ratio), and connection methods (such as hinges, joints, welding, or bolting). Three-dimensional mesh elements are then generated using professional finite element software (such as ANSYS and ABAQUS). A hybrid strategy of structured and unstructured meshing is employed to address the geometric characteristics of different components, improving computational accuracy in critical stress-bearing areas. Furthermore, the modeling process requires special consideration of the end-effector's structural extension, as its additional loads and connection stiffness significantly influence the overall system's modal characteristics and are crucial factors in subsequent excitation coupling analysis.

[0018] Secondly, based on the finite element model, multiple groups of operating conditions are set to simulate the posture and load state changes of the robotic arm in typical tasks. The working conditions should cover common loading postures (such as horizontal grasping, vertical delivery, angular rotation insertion, etc.), as well as different clamping weight conditions related to the end load. Under each group of working conditions, the corresponding boundary conditions (such as fixed constraints, joint degree of freedom definition) and mass loads are applied to the model, and modal analysis calculations are performed to extract the first several modal frequencies and their corresponding vibration mode distributions. By comparing and analyzing the modal results under different working conditions, it can be found that some structural components have frequency drift and vibration mode concentration in specific motion configurations, which is manifested as the modal frequency falling into a relatively concentrated frequency band area and exciting similar local structural resonances under specific configurations. These frequency bands are potential frequency sensitive areas.

[0019] Third, the modal analysis results obtained under different operating conditions are subjected to data aggregation and statistical processing. A frequency band clustering algorithm is used to aggregate the modal points with high-frequency overlap in the frequency distribution to form an initial set of frequency-sensitive areas. Combining the distribution position of the modal vibration shape with the structural energy density data, the key structural units with frequent response concentration under multiple operating conditions are screened out, and the "dynamic response active section" related to trajectory control is further refined. On this basis, a frequency-sensitive area parameter set is established through indicators such as characteristic frequency intervals, structural position information, and response morphology to form a multi-dimensional digital model that describes the structural excitation risk. This parameter set will serve as an important reference standard for excitation frequency matching analysis in the subsequent path planning stage.

[0020] Finally, the frequency-sensitive area parameter set generated above is stored in the frequency characteristic database module in the control system, and the data interface is connected to the subsequent trajectory excitation sequence analysis module to realize real-time resonance risk identification during the path planning process. In the database, each frequency-sensitive parameter record should include fields such as the working condition number, key frequency band range, modal number, structural action area identification, energy concentration coefficient, etc., to facilitate the control algorithm to formulate dynamic avoidance strategies during path interpolation. In order to ensure the long-term adaptability of the data, the frequency-sensitive data can be periodically updated and corrected in combination with the structural health monitoring feedback of the manipulator in actual operation to enhance the robustness of the model to structural changes. Through the above four steps, this embodiment realizes the frequency response characteristic modeling and parameter extraction of the manipulator structure under high dynamic operating conditions, providing basic data support for subsequent trajectory resonance avoidance control.

[0021] The robot's planned running path is segmented and expanded into a time-domain excitation signal sequence. The excitation frequency characteristics of each trajectory segment in the sequence are calculated. The frequency characteristics are matched with the frequency-sensitive region parameter set to identify the trajectory segments that have the risk of exciting structural resonance. To identify and process potential resonance-excitation risk segments within the robot's planned path, we combine time-domain segmentation of path signals with frequency-domain feature analysis to construct a digital representation of the mechanical path excitation frequency. This process then extracts and identifies high-risk segments by matching them with a set of frequency-sensitive region parameters. This process involves four steps: First, according to the preset loading rhythm and operation process requirements, the complete loading operation path data of the robot is obtained. The path data can be derived from the output of the offline trajectory planning system or the online teaching system, and the data form includes multi-dimensional path parameters such as displacement sequence, velocity curve, acceleration instruction, etc. On this basis, the path is segmented uniformly or non-uniformly according to the time dimension to form path segments within multiple time interval windows. In each trajectory segment, the time series data related to the execution dynamics is extracted to construct a time domain excitation signal sequence representing the path characteristics, specifically including acceleration pulse sequence, velocity change rate, displacement change rate, etc. This step aims to convert discrete trajectory segments into signal inputs with dynamic stimulation characteristics, providing a data basis for subsequent frequency domain processing.

[0022] Secondly, for the above-mentioned time-domain excitation signal sequence, a frequency domain analysis method is used to extract its frequency characteristic indicators. In this embodiment, digital signal processing technologies such as fast Fourier transform (FFT) or short-time Fourier transform (STFT) are used to convert the acceleration or velocity signal in each trajectory segment into spectral information. By performing power spectral density analysis on the spectrum graph, indicators such as the main frequency, secondary frequency, frequency bandwidth and energy concentration of the excitation signal in each trajectory segment are identified. To enhance the resolution of the analysis, window functions and overlapping window algorithms can be introduced to refine and compensate for the frequency components of the boundary segments. The extraction of frequency features is not limited to single peak identification, but rather constructs a data set containing multi-dimensional descriptions such as the main frequency distribution, energy weight coefficient, and frequency change trend, thereby providing sufficient frequency domain characterization capabilities for subsequent matching analysis.

[0023] Third, based on the extracted frequency feature data, a segment-by-segment matching analysis is performed with a pre-established set of frequency-sensitive region parameters. In this step, the primary and secondary frequency distributions of all path trajectory segments are traversed and their overlap with the natural frequency band range in the sensitive frequency set is determined. If a frequency component in the trajectory segment falls within any frequency-sensitive interval and its frequency energy ratio exceeds a set threshold (e.g., a 30% power ratio), the trajectory segment is determined to have a structural excitation risk. The matching results not only output "resonance" in Boolean format but also mark each trajectory segment with metadata such as the resonance risk level, the possible excited structural location, and the corresponding modal number, forming a trajectory segment risk identification matrix. This matrix will be used in subsequent disturbance response prediction and path avoidance design.

[0024] Finally, the data of the trajectory segments identified as having resonance risks are archived and visualized, and a map-based identification mechanism for path resonance excitation is established. Combined with the spatial position information of the trajectory segments in the actual motion path of the manipulator, a time-path two-dimensional coordinate mapping diagram is constructed, and the trajectory segments with high resonance risks are presented in the form of color gradients or risk indices. On this basis, the data characteristics of the high-risk trajectory segments (such as excitation frequency, matching mode, and trigger condition number) are fed back to the control algorithm module as key inputs for path reconstruction, acceleration modulation, and operation strategy optimization. In addition, the analysis results can be linked with the operation data to establish path versions and resonance risk archives, providing historical basis and parameter references for path evaluation under different batches or different task conditions. Through the above four steps, this embodiment realizes the quantitative identification and information modeling of resonance risks in the manipulator path, improves the feedforward perception capability of the path control system for structural responses under high dynamic excitation conditions, and lays the foundation for dynamic trajectory optimization and system stability enhancement.

[0025] For the identified trajectory segments with resonance excitation risks, a time-domain disturbance response model is constructed to simulate the disturbance propagation path and response amplitude of the current trajectory segment during continuous operation. The offset of the trajectory disturbance on the terminal posture accuracy is calculated, and a resonance cumulative risk index is generated based on the response trend. In order to conduct an in-depth assessment of the identified trajectory segments with the risk of structural resonance excitation, and further determine the potential disturbance impact on the end-point posture control accuracy and the cumulative risk during continuous operation, a trajectory disturbance analysis method based on time-domain disturbance modeling and response simulation is proposed. A multi-order response model is established to quantify the structural excitation trend and system accuracy degradation risk of the manipulator under high-speed operation. The method includes the following four steps: First, based on the resonance risk trajectory segment identified in the previous stage, the key dynamic parameters of this trajectory segment during actual operation are extracted to construct a disturbance input signal. This disturbance input signal primarily includes control variables with dynamic drive characteristics, such as the acceleration excitation function, the slope change rate of the path interpolation curve, and the instantaneous current pulse sequence of the servo drive system. On this basis, a multi-body dynamics model is constructed for the multi-joint coupling system of the robotic arm. Combining structural modal information with material damping properties, a time-domain response method (such as the time-step integration method) is used to calculate the system's structural response process under different input excitation conditions. The model considers the connection flexibility between the end effector and the main arm structure, the inertial coupling effect, and the gaps or small nonlinear friction characteristics existing in the actual joints to ensure the accuracy and operability of the disturbance response prediction results.

[0026] Secondly, after the structural response model is established, a time-stepping simulation is performed for each high-risk trajectory segment, outputting key response variables during the path execution process, including the end-point posture drift, the structural strain distribution of each joint point, the end-point acceleration peak value, and the trend of the reverse torque change. By performing differential analysis on the simulation results, the spatial propagation path of the disturbance response can be obtained, that is, the transmission path of the disturbance from the excitation source point to the end of the structure and the main response concentration area can be determined. At the same time, multiple groups of working conditions are introduced for response comparison, and a mapping relationship between the disturbance response intensity and the path dynamic parameters is constructed. The sensitivity of path parameter changes to the dynamic stability of the structure is clarified, providing decision support for subsequent optimization control.

[0027] Third, based on the time-domain simulation output, a cumulative indicator system is constructed to characterize the degree of resonant excitation risk. In this embodiment, the "Trajectory Disturbance Risk Factor (TD-RI)" is defined as the primary indicator. Its calculation model comprehensively considers factors such as terminal drift amplitude, frequency response overlap, response delay time, and the persistence intensity of the disturbance. By comparing the response superposition trends of each trajectory segment under continuous execution conditions, it is determined whether there is a phenomenon of gradually amplifying the resonant response, that is, whether the trajectory segment has a "cumulative excitation tendency." At the same time, the "Joint Stress Risk Value (JS-RV)" and "Servo Response Deviation (SC-DV)" are constructed as supplementary indicators to identify the fatigue trend of the joint structure and the degree of nonlinear response of the control signal. All risk indicators are standardized to a unified dimension and can be visually tracked along the time axis for dynamic call-up during the subsequent path optimization process.

[0028] Finally, the trajectory disturbance response results and the resonance cumulative risk index generated above are integrated into the intelligent analysis module of the trajectory scheduling system to construct a trajectory segment risk profile, which is compared with historical data to achieve adaptive risk weight allocation. During path planning or operation strategy adjustment, when the system detects that the TD-RI or JS-RV value of a certain trajectory segment exceeds the set threshold, the trajectory segment is automatically marked as a "path intervention priority segment" for the path reconstruction module to call. At the same time, based on the evolution trend of the structural state in historical operation, the risk index of the current trajectory segment can be predictively extrapolated to determine the further stability deterioration trend that may be triggered in high-frequency operation. Through the above four steps, this embodiment realizes the refined modeling and response visualization evaluation of the dynamic disturbance risk of the manipulator trajectory segment, providing basic support for the active avoidance of structural resonance problems and the guarantee of operational stability.

[0029] Based on the resonance cumulative risk index, dynamic trajectory path reconstruction is performed, and the acceleration pulse sequence of the target trajectory is adjusted using a nonlinear adjustment function to generate a low-frequency coupling path sequence that avoids the natural frequency of the structure. To address the dynamic resonance problem caused by the coupling between the path excitation frequency and the structural natural frequency during high-speed continuous operation of the manipulator, a dynamic trajectory reconstruction method is proposed based on the resonance cumulative risk index identified in the early stage. By nonlinearly adjusting the acceleration pulse sequence, the frequency-sensitive range of the structure is effectively avoided, thereby generating a new trajectory path that matches the low-frequency coupling characteristics of the structure. The method includes the following four steps: First, based on the set of resonance cumulative risk indicators output by the trajectory perturbation analysis module, the dynamic response characteristics of each trajectory segment are prioritized to identify high-risk trajectory segments requiring path reconstruction. This ranking process utilizes multiple metrics, including "trajectory perturbation risk factor," "end-position deviation trend," and "joint stress amplitude cumulative coefficient," to establish a path risk assessment matrix. The selected trajectory segment serves as the target for optimization, and its original path parameters, including the target position point set, trajectory interpolation velocity, acceleration curve, and pulse update period, are extracted. This provides a parameter basis for reconstruction design while maintaining a consistent task action sequence. This step aims to establish the path optimization target space and clarify the mapping relationship between control variables and evaluation criteria.

[0030] Secondly, for the above-mentioned trajectory segment, a family of nonlinear adjustment functions is constructed to adjust the changing trend of its acceleration pulse sequence. The nonlinear adjustment function may include an exponential function, a sinusoidal modulation function, a hyperbolic tangent function, a polynomial transition function, etc. The specific function form is determined according to the target frequency avoidance bandwidth and the tolerance to dynamic response changes. In the process of reconstructing the acceleration curve, the "frequency amplitude transfer" strategy is adopted, that is, when the original trajectory excitation frequency falls near the natural frequency of the structure, the acceleration rising edge slope, the maximum acceleration amplitude and the duration are nonlinearly scaled so that the main frequency of the actual excitation signal migrates to a lower frequency band, thereby avoiding the structural modal response interval. In order to maintain the overall geometric accuracy of the trajectory, a multi-point interpolation boundary condition control function convergence boundary is introduced to ensure that the displacement time integral result is equivalent to the original path, avoiding target dislocation caused by trajectory deformation.

[0031] Third, after the nonlinear trajectory reconstruction function is applied, a frequency domain analysis is performed, and the reconstructed path excitation signal is Fourier transformed to verify whether its dominant frequency and energy distribution have successfully avoided the structural frequency-sensitive areas. In this step, the "frequency coupling suppression coefficient" is introduced as a judgment metric. This coefficient is defined as the proportion of energy in the target excitation spectrum that falls within the structural natural frequency range. If this proportion is below a preset threshold (e.g., 5%), the trajectory reconstruction is considered to be satisfactory. For trajectory segments that do not meet the avoidance requirements, the modulation function parameters are adjusted or the reconstruction process is repeated using another modulation function, forming an iterative path optimization process. This process can adopt intelligent optimization strategies such as genetic algorithms, particle swarm optimization, or dynamic programming to improve the efficiency and robustness of path generation.

[0032] Finally, the low-frequency coupling path sequence that has passed the frequency suppression verification is embedded in the trajectory scheduling module of the manipulator control system, and an error comparison and scheduling test is performed with the original path. During the test, the system dynamically compensates the downstream process synchronization mechanism based on the execution time difference introduced by the trajectory reconstruction to maintain the stability of the production rhythm. At the same time, the operating data such as the execution current fluctuation, joint response smoothness and end positioning error of the reconstructed path during the servo system control process are recorded and fed back to the optimization module for subsequent trajectory library updates. The reconstructed trajectory sequence will also be stored in the path version database to form a path candidate set for different structural resonance modes for the system to call and match under different task conditions. Through the above four steps, this embodiment realizes the dynamic reconstruction of the path based on the frequency avoidance mechanism. Without changing the original task logic, it reduces the risk of path excitation triggering structural resonance and improves the structural stability and trajectory control accuracy of the system under continuous operation.

[0033] The low-frequency coupling path sequence is input into the manipulator scheduling module. A dynamic trajectory verification mechanism is set up in the early stage of control system operation. The structural response is tested by micro-injection of trajectory disturbances, and the path control parameters are corrected in real time based on the response data. To verify the structural response of the low-frequency coupling path sequence reconstructed using the frequency avoidance method during actual operation and further improve the effectiveness and individualized matching of path control, a dynamic verification mechanism combining trajectory perturbation microinjection with real-time structural response feedback is proposed. This mechanism is executed during the initial operation of the control system. By embedding trajectory perturbation test points and monitoring the corresponding structural responses, it achieves online correction and optimization of path control parameters. This implementation method mainly includes the following four steps: First, the low-frequency coupling path sequence generated by the frequency coupling reconstruction algorithm is imported into the trajectory scheduling module of the robot control system and dynamically bound to the existing operation rhythm control logic and workstation coordination mechanism. During the path import process, the system performs timestamp mapping and buffer queue management on the trajectory execution points to ensure that the new path sequence can seamlessly replace the original path during execution without affecting the downstream operation sequence. The path data includes multi-dimensional instruction content, such as the target position point set, velocity vector, acceleration pulse sequence, and pulse emission interval, and is loaded into the scheduling core logic in the form of task blocks. This step ensures the parameter consistency and data integrity of the path between the program layer and the physical execution layer.

[0034] Secondly, during the control system initialization phase—that is, before the robot actually begins operating but enters formal rhythmic mode—a trajectory perturbation micro-injection test is performed. This process simulates dynamic response fluctuation scenarios by presetting micro-perturbation signals in specific trajectory segments. For example, this involves perturbing the amplitude of the acceleration pulse sequence by ±5%, introducing very small changes in velocity slope, or adjusting the time intervals between path nodes. The perturbation must be designed to ensure that it falls within the system's safety margin, neither causing destructive excitation to the structure nor interfering with the main path task logic. The perturbation signal is injected sequentially between multiple critical path nodes, and response data, including end-position drift, acceleration trend, and joint drive current fluctuations, is collected in conjunction with a posture detection system (such as a high-precision encoder, laser interferometer, and end-position vision measurement system).

[0035] Third, the control system analyzes the deviation between the response trend and the expected trajectory behavior in real time based on the structural response data collected after trajectory disturbance injection. To this end, this embodiment constructs a trajectory response evaluation index system covering core parameters such as "end attitude response error rate," "structural response delay time," and "post-disturbance path deviation correction rate," and maps these parameters to the path control parameters for analysis. When the disturbance response error detected in a certain segment exceeds a threshold (for example, a position error exceeding 0.3mm or a current fluctuation exceeding a set percentage), the system determines that the path segment still has a risk of resonant coupling or a parameter mismatch. At this point, the scheduling module automatically triggers the path correction mechanism and, based on the response results, makes detailed adjustments to the control parameters for that trajectory segment. This can include fine-tuning the acceleration pulse rising edge slope, delaying the execution time between specific points, or reducing the servo response sensitivity, thereby achieving a more consistent dynamic match between the control system and the actual structural behavior.

[0036] Finally, the system relabels the path after the micro-injection test and parameter correction as a "verification path" and performs a full simulation verification before the robot officially enters the operation beat mode to ensure that the corrected path has the characteristics of minimizing structural excitation without affecting the production rhythm. If all response data of the verification path in the simulation are lower than the control tolerance limit, the path is stored in the path version database and marked as a "structural correction adaptation path"; otherwise, it automatically rolls back to the previous version path and prompts the control personnel to make further intervention adjustments. This mechanism also has the ability to continuously optimize during operation. After the system is started, it can re-inject disturbances at fixed intervals (such as every 500 beats) to perform structural health checks to adapt to possible slight changes in the structural state over time. Through the above four steps, this embodiment realizes the dynamic structural response verification and path parameter accuracy correction after trajectory reconstruction, effectively improving the robustness and accuracy consistency of the path control system in the actual structural excitation environment.

[0037] During the operation of the control system, periodic structural response verification operations are performed to collect data on fluctuations in end-point posture accuracy, path execution residuals, and changes in mechanical joint drive currents to evaluate operational stability. If abnormal responses are detected, path re-optimization operations are triggered. To ensure the long-term structural response stability of the robot during continuous loading, a periodic structural response verification mechanism is proposed. This mechanism is used to regularly collect key response data during the control system operation cycle, comprehensively evaluate the system operation status, and automatically trigger path re-optimization operations when an anomaly is detected. The specific steps are as follows: During the operation of the control system, the structural response data is periodically collected at fixed working cycles (for example, every 500 path execution cycles). The key parameters collected include the end-point pose accuracy fluctuation, path execution residual, and the rate of change of the mechanical joint drive current. The structural stability monitoring factor is calculated based on the acquired parameters. The calculation expression is as follows: , where It is the average offset between the actual position and the theoretical position of the end of the robot during the execution of the path. The unit is millimeter. The real-time posture acquisition is performed through high-precision measurement equipment such as encoders, laser interferometers, and three-dimensional vision systems. The average offset of all sampling points in each execution cycle is calculated. is the cumulative error between the actual execution trajectory and the target trajectory, It is the comparison ratio of the current fluctuation in the previous cycle to the previous cycle, expressed in percentage. 、 as well as They are 、 as well as The weight coefficient of Through weight configuration, we can dynamically focus on different system indicators, balance the trade-offs of "structural accuracy", "control trajectory consistency" and "energy consumption disturbance", and adjust the weights of the three parameters in the overall evaluation. It is a structural stability monitoring factor, which is an indicator obtained by comprehensive weighted calculation of terminal posture fluctuation, path execution residual and servo current change. It reflects the overall structural stability and control coordination of the system in the current cycle; By calculation , to achieve quantitative characterization of the current system operation stability.

[0038] Obtaining the structural stability monitoring factor of the current cycle Then, combine it with the previous The mean values ​​within a period are compared to determine whether there is an abnormal operating status. The difference is expressed in the form of relative deviation. The calculation formula is: , where It is before The structural stability monitoring factor of the cycle refers to the past continuous The average value of the stability factor calculated within the operating cycle represents the stability reference baseline of the system under normal operating conditions. The stability deviation is an indicator that measures the relative change between the current operating state and the historical state. It is defined in the form of a ratio, making it universal for different system scales. It is used to reflect the degree of difference between the current state and the historical state. When its value is too large, it indicates that the system has experienced obvious disturbances, deviations or structural abnormalities. It is a key variable for determining whether to trigger re-optimization. If the calculated If it is greater than the set trigger threshold (for example, 0.15), it is determined that the current control system state has abnormal fluctuations, which may be caused by structural resonance caused by path excitation or micro-drift of the control system. The re-optimization trigger flag set by the control system will be set to 1: , where The path reoptimization trigger threshold is a preset trigger boundary value used to determine at what degree of deviation it should be considered an "abnormal state" and thus activate the trajectory optimization process. The value range is expressed in decimals (for example, 0.15 represents a 15% deviation). The flag for triggering path reoptimization is a control variable used by the system to determine whether the path needs to be reoptimized. Its value is Boolean (0 or 1), indicating whether the path correction operation is triggered. It serves as a communication signal between the system path scheduling module and the optimization module, directly controlling whether the trajectory perturbation reconstruction algorithm is executed, avoiding unnecessary path switching, and ensuring the continuity and stability of system operation. This flag will serve as a starting signal for the system path re-optimization mechanism, ensuring that adaptive adjustments are only made when a decrease in structural stability is actually detected, avoiding frequent and ineffective optimization operations.

[0039] When the path reoptimization trigger flag When activated, the control system will execute the path parameter disturbance correction process. The correction process is based on the original acceleration pulse of the current path, and performs dynamic correction to reduce the degree of excitation frequency coupling. The new trajectory acceleration curve is generated according to the following formula: , where The target acceleration pulse of the original path represents the acceleration pulse of the robot at each moment during the trajectory execution. The target acceleration value under is the basic motion instruction output by the trajectory control system according to the interpolation planning algorithm. As the core input of the dynamic characteristics of the path, it determines the acceleration and deceleration behavior of the manipulator in each segment of the path. Its shape determines the trajectory excitation frequency distribution. It is the disturbance control gain coefficient, which controls the sensitivity and amplitude adjustment of the path acceleration correction and determines the intensity of the response to the abnormal response deviation of the system. The value range is between 0.3 and 0.7. The specific value is set according to the system rigidity, response inertia and the anti-disturbance ability of the control system. The larger the value, the more sensitive the system is to abnormal changes and the stronger the correction. The smaller the value, the more stable and conservative the system is and the smaller the adjustment amplitude is. It plays the role of a dynamic adjustment force limiter to prevent excessive correction from causing new vibration or control hysteresis. After the new track is generated, save it as a version , with the original path version For comparative testing, the new path will be collected again after actual operation 、 as well as value, calculate the new , compared with the old version, if the new stability index calculated after the path execution meets the following conditions: , the new path is considered to significantly improve the system response state, replaces the old path and enters production use; if the condition is not met, roll back to the old path and record the deviation data for subsequent optimization iterations, where It is a structural stability monitoring factor, which represents the comprehensive stability index calculated by the structural response data (such as position offset, current fluctuation, and path residual) collected by the system in the new cycle after the optimized path is adopted and executed. The larger the value, the higher the degree of system instability. is the structural stability monitoring factor under the original path, which represents the structural stability index collected and calculated when using the old path (i.e., the trajectory without frequency avoidance and disturbance suppression optimization processing).

[0040] Through the above three steps, this implementation method realizes a closed-loop control mechanism of periodic evaluation of structural response, abnormal trigger identification and dynamic re-optimization of path parameters, which significantly improves the stability and control accuracy of the manipulator system under long-term high-speed operation.

[0041] By constructing a finite element simulation model of the robotic arm structure and combining segmented analysis of the path excitation frequency with frequency-sensitive region matching, this invention can proactively identify trajectory segments with a risk of resonance excitation before the manipulator actually executes the path, thereby preventing the coupling of path excitation with the structural natural frequency. Compared to traditional control methods that rely on empirical settings or post-correction, this invention achieves feedforward identification and early intervention of dynamic structural responses, significantly improving the trajectory control system's dynamic adaptability and ability to prevent structural excitation issues. It effectively reduces system risks caused by resonance, such as structural fatigue, end-point precision drift, and joint control errors.

[0042] The present invention forms an adaptive closed-loop path optimization system through dynamic trajectory path reconstruction, disturbance micro-injection testing and periodic response verification mechanism. When the robot arm has trajectory execution deviation caused by structural aging, load changes or environmental disturbances in actual operation, the system can collect data such as end posture, path residual, current fluctuation in real time, and automatically trigger path re-optimization operation based on the structural stability monitoring factor. This dynamic path adjustment mechanism that integrates structural response feedback not only improves the stability and robustness of the control system in long-term continuous operation, but also enables the entire production line to have higher self-recovery capabilities and intelligent operation and maintenance level. It is particularly suitable for automated loading scenarios with high beats, long cycles and high precision positioning requirements.

[0043] A continuous loading manipulator control system includes a structural modal analysis module, a trajectory excitation feature recognition module, a disturbance response modeling module, a path reconstruction and optimization module, a trajectory dynamic verification module, and a structural stability monitoring and re-optimization module: The structural modal analysis module builds finite element simulation models for each joint and end-effector structure of the robotic arm, obtains the natural frequency distribution data of the structure under different load states and posture conditions, and generates a set of frequency-sensitive region parameters based on the data; The trajectory excitation feature recognition module breaks down the manipulator's planned path into segments and converts them into a time-domain excitation signal sequence. It then calculates the excitation frequency characteristics of each trajectory segment in the sequence, matches these frequency characteristics with a set of frequency-sensitive region parameters, and identifies trajectory segments that pose a risk of exciting structural resonance. The disturbance response modeling module constructs a time-domain disturbance response model for identified trajectory segments with resonance excitation risks. It simulates the disturbance propagation path and response amplitude of the current trajectory segment during continuous operation, calculates the offset of the trajectory disturbance on the terminal posture accuracy, and generates a resonance cumulative risk index based on the response trend. The path reconstruction optimization module performs dynamic trajectory path reconstruction based on the resonance cumulative risk index and uses a nonlinear adjustment function to adjust the acceleration pulse sequence of the target trajectory to generate a low-frequency coupling path sequence that avoids the natural frequency of the structure. The trajectory dynamic verification module inputs the low-frequency coupling path sequence into the manipulator scheduling module, sets up a dynamic trajectory verification mechanism in the early stage of control system operation, tests the structural response by micro-injecting trajectory disturbances, and performs real-time correction of path control parameters based on the response data; The structural stability monitoring and re-optimization module performs periodic structural response verification operations during the operation of the control system, collects data on end-position accuracy fluctuations, path execution residuals, and mechanical joint drive current changes, evaluates the stability of the operating state, and triggers path re-optimization operations if it is determined that there is a response abnormality.

[0044] A continuous loading robot control method provided in an embodiment of the present invention is realized by the above-mentioned continuous loading robot control system. The specific method and process of a continuous loading robot control system are detailed in the embodiment of the above-mentioned continuous loading robot control method, which will not be repeated here.

[0045] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. It is apparent to those skilled in the art that various changes, modifications, substitutions, and variations to these embodiments may be made without departing from the principles and spirit of the present invention, and these changes and modifications still fall within the scope of protection of the present invention.

Claims

1. A continuous feeding robot control method, characterized in that: The following steps are involved: Construct finite element simulation models for each joint and end-effector structure of the robotic arm, obtain the natural frequency distribution data of the structure under different load states and posture conditions, and generate a set of frequency-sensitive region parameters based on the data; The robot's planned running path is segmented and expanded into a time-domain excitation signal sequence. The excitation frequency characteristics of each trajectory segment in the sequence are calculated. The frequency characteristics are matched with the frequency-sensitive region parameter set to identify the trajectory segments that have the risk of exciting structural resonance. For the identified trajectory segments with resonance excitation risks, a time-domain disturbance response model is constructed to simulate the disturbance propagation path and response amplitude of the current trajectory segment during continuous operation. The offset of the trajectory disturbance on the terminal posture accuracy is calculated, and a resonance cumulative risk index is generated based on the response trend. Based on the resonance cumulative risk index, dynamic trajectory path reconstruction is performed, and the acceleration pulse sequence of the target trajectory is adjusted using a nonlinear adjustment function to generate a low-frequency coupling path sequence that avoids the natural frequency of the structure. The low-frequency coupling path sequence is input into the manipulator scheduling module. A dynamic trajectory verification mechanism is set up in the early stage of control system operation. The structural response is tested by micro-injection of trajectory disturbances, and the path control parameters are corrected in real time based on the response data. During the operation of the control system, periodic structural response verification operations are performed to collect data on the end-point posture accuracy fluctuation, path execution residuals, and mechanical joint drive current changes to evaluate the stability of the operating state. If it is determined that there is a response abnormality, the path re-optimization operation is triggered.

2. A continuous feeding manipulator control method according to claim 1, characterized in that: The specific steps for building a finite element simulation model of each joint and end-effector structure of the robotic arm are as follows: A finite element model is built based on the structural components of the robotic arm. The model includes the base, multiple joints, link units, and end fixtures, and meshes are divided according to size, material properties, and connection methods. Set up multiple groups of posture and load conditions, apply boundary conditions and loads to each group of conditions, perform modal analysis, and extract modal frequencies and vibration mode distributions; The frequency band clustering method is used to statistically process the modal data, extract the active response sections, and construct the frequency sensitive area parameter set; The frequency-sensitive area parameter set is stored in the frequency feature database module, and trajectory resonance risk identification and dynamic update are supported.

3. A continuous feeding manipulator control method according to claim 1, characterized in that: The specific steps of segmenting the manipulator's planned running path into a time domain excitation signal sequence and identifying the trajectory segments that induce structural resonance are as follows: Obtain the robot's loading path data once, and construct a time-domain excitation signal sequence containing acceleration, velocity, and displacement changes in segments according to the time dimension; Perform frequency domain analysis on each trajectory segment to extract the excitation frequency characteristics, including the main frequency, secondary frequency, frequency energy distribution and frequency change trend; Match the frequency features with the frequency-sensitive region parameter set to identify trajectory segments with overlapping frequencies and energy proportions exceeding a threshold as those with structural excitation risks. Data archiving and visualization of the identified trajectory segments are performed, and a risk identification matrix is ​​generated for path reconstruction.

4. A continuous feeding manipulator control method according to claim 1, characterized in that: The specific steps for constructing a time-domain disturbance response model for the identified trajectory segments with resonance excitation risks are as follows: Extract the dynamic control parameters of the resonance risk trajectory segment, construct the acceleration excitation function and current pulse as the disturbance input, establish the multi-body dynamic model and perform time domain response calculation; Perform time-stepping simulation to output terminal pose drift, structural strain and torque changes, analyze disturbance propagation paths, and construct a mapping relationship between path parameters and structural responses; Based on the simulation results, the trajectory disturbance risk factor, joint stress risk value, and servo response deviation are calculated and standardized into a unified risk index. Integrate risk indicators into the trajectory scheduling system to generate path risk profiles and mark priority segments for path intervention.

5. The method for controlling a continuous loading robot according to claim 1, characterized in that: The specific steps for performing dynamic trajectory path reconstruction based on the resonance cumulative risk index are as follows: The trajectory segments are sorted according to the trajectory disturbance risk factor and the cumulative coefficient of joint stress amplitude, and the high-risk trajectory segments and their original path parameters are extracted as reconstruction targets; A nonlinear regulation function is constructed to adjust the acceleration pulse sequence, and the excitation main frequency is shifted to outside the natural frequency of the structure through frequency amplitude transfer. Perform frequency domain analysis and calculate the frequency coupling suppression coefficient. If the energy ratio is lower than the threshold, the reconstruction is considered to meet the standard. If it does not meet the standard, re-optimize; The reconstructed path is embedded in the scheduling module and the servo execution response data is recorded, and the path version database is updated.

6. A continuous feeding manipulator control method according to claim 1, characterized in that: The steps of inputting the low-frequency coupling path sequence into the manipulator scheduling module and setting up the dynamic trajectory verification mechanism include: Import the low-frequency coupling path sequence into the trajectory scheduling module, complete timestamp mapping and buffer queue management, and load it into the task block execution logic; During the initialization phase, a small-amplitude disturbance signal is injected into multiple path nodes and the terminal position drift, acceleration change, and current fluctuation response data are collected; Triggering the path parameter correction mechanism based on the response deviation to perform acceleration slope adjustment, inter-node time delay, or servo response sensitivity reduction operations; The verified path mark is archived, and after simulation verification, it is determined whether to write it into the path version database as a structural correction adaptation path.

7. A continuous feeding manipulator control method according to claim 1, characterized in that: To ensure the long-term structural response stability of the robot during continuous loading, a periodic structural response verification mechanism is proposed. This mechanism is used to regularly collect key response data during the control system operation cycle, comprehensively evaluate the system operation status, and automatically trigger path re-optimization operations when an anomaly is detected. The specific steps are as follows: During the operation of the control system, the periodic acquisition of structural response data is performed in units of a fixed working cycle. The key parameters collected include the end-point pose accuracy fluctuation, the path execution residual, and the rate of change of the mechanical joint drive current. The structural stability monitoring factor is calculated based on the acquired parameters. The calculation expression is as follows: , where It is the average offset between the actual position and theoretical position of the end of the robot during the execution of the path. is the cumulative error between the actual execution trajectory and the target trajectory, is the ratio of the current fluctuation in the previous cycle to that in the previous cycle, 、 as well as They are 、 as well as The weight coefficient of is the structural stability monitoring factor; Obtaining the structural stability monitoring factor of the current cycle Then, combine it with the previous Compare the mean values ​​within a period to determine whether there is an abnormal operating status. The calculation formula is: , where It is before Structural stability monitoring factor of the cycle, is the stability offset; If the calculated If it is greater than the set trigger threshold, it is determined that the current control system state has abnormal fluctuations, and the re-optimization trigger flag set by the control system will be set to 1: , where is the path reoptimization trigger threshold, It is the flag bit that triggers path re-optimization; When the path reoptimization trigger flag When activated, the control system will execute the path parameter disturbance correction process to generate a corrected new trajectory acceleration curve, as shown in the following formula: , where The target acceleration pulse of the original path, is the disturbance control gain coefficient; After the new track is generated, save it as a version , with the original path version For comparative testing, the new path will be collected again after actual operation 、 as well as value, calculate the new , compared with the old version, if the new stability index calculated after the path execution meets the following conditions: , the new path is considered to significantly improve the system response state, replaces the old path and enters production use; if the condition is not met, roll back to the old path and record the deviation data for subsequent optimization iterations, where is the structural stability monitoring factor, It is the structural stability monitoring factor under the original path.

8. A continuous feeding manipulator control system, used to implement a continuous feeding manipulator control method according to any one of claims 1 to 7, characterized in that: It includes structural modal analysis module, trajectory excitation feature identification module, disturbance response modeling module, path reconstruction optimization module, trajectory dynamic verification module and structural stability monitoring and re-optimization module: The structural modal analysis module builds finite element simulation models for each joint and end-effector structure of the robotic arm, obtains the natural frequency distribution data of the structure under different load states and posture conditions, and generates a set of frequency-sensitive region parameters based on the data; The trajectory excitation feature recognition module breaks down the manipulator's planned path into segments and converts them into a time-domain excitation signal sequence. It then calculates the excitation frequency characteristics of each trajectory segment in the sequence, matches these frequency characteristics with a set of frequency-sensitive region parameters, and identifies trajectory segments that pose a risk of exciting structural resonance. The disturbance response modeling module constructs a time-domain disturbance response model for identified trajectory segments with resonance excitation risks. It simulates the disturbance propagation path and response amplitude of the current trajectory segment during continuous operation, calculates the offset of the trajectory disturbance on the terminal posture accuracy, and generates a resonance cumulative risk index based on the response trend. The path reconstruction optimization module performs dynamic trajectory path reconstruction based on the resonance cumulative risk index and uses a nonlinear adjustment function to adjust the acceleration pulse sequence of the target trajectory to generate a low-frequency coupling path sequence that avoids the natural frequency of the structure. The trajectory dynamic verification module inputs the low-frequency coupling path sequence into the manipulator scheduling module, sets up a dynamic trajectory verification mechanism in the early stage of control system operation, tests the structural response by micro-injecting trajectory disturbances, and performs real-time correction of path control parameters based on the response data; The structural stability monitoring and re-optimization module performs periodic structural response verification operations during the operation of the control system, collects data on end-position accuracy fluctuations, path execution residuals, and mechanical joint drive current changes, evaluates the stability of the operating state, and triggers path re-optimization operations if it is determined that there is a response abnormality.

Images