A control method and device for a high-precision multi-joint linkage mechanical arm

Abstract

This invention discloses a control method and device for a high-precision multi-joint linkage robotic arm, relating to the field of robot control technology. The method includes the following steps: during the control process of the multi-joint linkage robotic arm, multi-source feedback information is collected and uniformly time-stamped. Alignment processing with the same sampling rhythm is performed using the uniform time stamp to form an aligned time sequence as the input sequence for subsequent fluctuation comparison. This invention maintains the continuity of the angle sequence in complex environments by performing time alignment, feature extraction, jump localization, and interference window adjustment on the multi-source feedback information, thereby improving motion stability and end-effector accuracy. Simultaneously, by adjusting the participation ratio of feedback information and delaying the entry of high-risk information into the fusion rhythm, the jump phase is buffered, and angle changes are smoothed through time misalignment, reducing the risk of accidental shutdown and enhancing the reliability of linkage operations.

Description

Technical Field

[0001] This invention relates to the field of robot control technology, specifically to a control method and device for a high-precision multi-joint linkage robotic arm. Background Technology

[0002] High-precision multi-joint linkage robotic arm control is one of the key technologies in the intelligent manufacturing equipment industry. It refers to the coordinated movement of all joints along a single trajectory by uniformly planning the angle, speed, and timing changes of each joint, given that the robotic arm has multiple independently rotating joints. This generates a continuous, repeatable, and jitter-free end-effector trajectory in space. The core of this technology lies in handling the coupling relationships between multiple joints. Based on kinematic and dynamic solutions, the target pose sequence is determined, and deviations, delays, and disturbances are corrected in real time during the driving process. This ensures that each joint's movement precisely follows the planned rhythm, ultimately allowing the robotic arm's end effector to accurately reach the desired position and posture. This supports applications in the intelligent manufacturing equipment industry such as complex path tracking, precision assembly, and high-speed operations.

[0003] The existing technology has the following shortcomings:

[0004] In existing technologies, multi-joint robotic arms typically rely on the synchronous convergence of various sensor data, such as angle encoders, acceleration signals, torque feedback, and current changes, to obtain real-time estimates of joint posture. However, in complex operating environments, these heterogeneous signals are prone to frequency overlap, phase disturbances, and synchronous noise increases within a very short period, leading to misinterpretation of the fused joint angle results as sudden jumps. These jumps are not actual joint movements but false indications caused by instantaneous mixing, easily misleading the system into believing that the joint has undergone high-speed rotation or abnormal twisting. Since trajectory planning requires real-time calculation of continuous joint angle sequences, misjudgment of angles directly disrupts trajectory following logic, causing end-effector posture deviation, disordered movement rhythm, or even triggering an emergency stop.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a control method and device for a high-precision multi-joint linkage robotic arm to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a control method for a high-precision multi-joint linkage robotic arm, comprising the following steps:

[0008] During the control of a multi-joint linkage robotic arm, multi-source feedback information is collected and uniformly time-stamped. The uniform time-stamped information is then used to perform alignment processing with the same sampling rhythm to form an aligned time sequence, which serves as the input sequence for subsequent fluctuation comparison.

[0009] The amplitude and phase changes of the signal in the input sequence are compared, short-time resonance segments are extracted between adjacent time points, mixing feature trajectories are generated, and the mixing feature trajectories are written to the end of the input sequence to form an extended sequence containing mixing features.

[0010] The continuous update trend of joint angles is detected along the mixing feature trajectory in the extended sequence, and the angle jump segment is located by combining the tail inertial change limit of the mixing feature trajectory, and a jump suspect list containing the jump time position is generated.

[0011] By reviewing the multi-source signal combination patterns in the extended sequence based on the suspected jump list, comparing the jump time position with the actual motion change trend in the extended sequence, separating pseudo-jump segments, and generating a draft interference window;

[0012] Based on the interference window draft, the fusion rhythm in the extended sequence is dynamically adjusted. Within the interference window, the participation ratio of multi-source feedback information is adjusted and the time when high-risk feedback information enters the fusion rhythm is delayed. Local time misalignment is introduced to smooth the angle sequence, so that the angle sequence avoids the jump input caused by instantaneous mixing.

[0013] Preferably, the input sequence formation steps are as follows:

[0014] During the control of a multi-joint linkage robotic arm, multi-source feedback information is collected and uniformly time-marked, forming a data set with consistent time points from the uniform time-marked data.

[0015] In a dataset with consistent time points, the sampling rhythm of different sensors is adjusted, and data with different sampling rhythms are integrated into a time series with a unified sampling rhythm.

[0016] In a time series with a uniform sampling rhythm, an aligned time series containing multiple types of feedback information is generated. The amplitude and phase changes in the aligned time series are compared and analyzed with time points as references.

[0017] An input sequence for fluctuation comparison is formed in the aligned time series of the comparative analysis. The input sequence is used as the basis for continuous joint angle updates and participates in the subsequent signal analysis process.

[0018] Preferably, the extended sequence formation steps are as follows:

[0019] The signal amplitude changes at each time point in the input sequence are compared, and the signal phase changes at the same time point are compared accordingly to form a joint relationship between amplitude changes and phase changes.

[0020] In the joint change relationship, identify the time interval where amplitude change and phase change fluctuate simultaneously between adjacent time points, and extract short-time resonance segments within this time interval;

[0021] The short-time resonance segments are arranged in chronological order to form a mixing characteristic trajectory, so that the mixing characteristic trajectory reflects the dynamic change characteristics generated by the interaction of multiple source signals.

[0022] The mixing feature trajectory is written to the end of the input sequence to form an extended sequence containing the mixing feature trajectory, which is then used in subsequent joint angle change analysis.

[0023] Preferably, the extraction of short-time resonance segments is based on the time interval in which amplitude and phase changes in the input sequence exhibit synchronous fluctuations at adjacent time points. The mixing feature trajectory records the joint change process of multi-source signals within this time interval and serves as the time correlation information in the extended sequence to reflect abnormal fluctuation characteristics.

[0024] Preferably, the steps for generating the suspected jump list are as follows:

[0025] The trend of joint angle change is continuously updated by following the mixing feature trajectory in the extended sequence, thus identifying the trajectory of joint angle change in the time series.

[0026] In the continuously updated trend detection results, the inertial change of the tail segment of the mixing feature trajectory is analyzed, and the inertial change limit corresponding to the angle change is determined.

[0027] Under the constraint of inertial change limit, identify the time interval of angle change exceeding the inertial change limit and locate the corresponding angle jump segment;

[0028] The time positions corresponding to the angle jump segments are recorded as jump time positions and compiled into a list of suspected jumps for use in subsequent sequence analysis.

[0029] Preferably, the jump time position in the jump suspect list is formed by marking the start and end time points of the angle jump segment in the extended sequence and arranging them in chronological order to limit the continuous time interval in which the angle change anomaly occurs.

[0030] Preferably, the steps for generating the interference window draft are as follows:

[0031] By reviewing the combination patterns of multi-source signals in the extended sequence around the jump time positions in the jump suspect list, the signal change trend within the corresponding time interval can be reproduced.

[0032] In the multi-source signal combination pattern obtained by review, the jump time position is compared with the actual motion change trend in the extended sequence to identify the angle change segment that is inconsistent with the motion change trend.

[0033] In the comparison results of real motion change trends, inconsistent angle change segments are separated to form a set of pseudo-jump segments.

[0034] An interference window draft is generated based on the distribution of the pseudo-jump fragment set on the time axis, and the morphological characteristics of the multi-source signal combination within the corresponding time interval are recorded.

[0035] Preferably, the interference window draft is divided on the time axis according to the continuous distribution of pseudo-jump segments, and the corresponding multi-source signal combination morphological characteristics are associated within each interference window to distinguish between the jump concentration interval and the normal change interval, and to limit the time range in the extended sequence that needs to participate in the fusion rhythm adjustment.

[0036] Preferably, the dynamic adjustment steps for the fusion rhythm are as follows:

[0037] Based on the draft interference window, the time interval corresponding to the interference window is determined in the extended sequence, and the multi-source feedback information is processed by time axis mapping within the time interval;

[0038] In the extended sequence of time-axis mapping processing, the participation ratio of multi-source feedback information is adjusted to form a signal combination form that conforms to the characteristics of the interference window;

[0039] In the signal combination pattern involved in the ratio adjustment, the time when high-risk feedback information enters the fusion rhythm is delayed, thereby changing the time position of high-risk feedback information in the extended sequence.

[0040] Local time misalignment is introduced into the extended sequence of delayed processing to smooth angle changes and form an angle sequence that avoids instantaneous mixing jump input.

[0041] A control device for a high-precision multi-joint linkage robotic arm includes a signal acquisition and alignment module, a feature extraction and expansion module, a jump detection module, a signal comparison and separation module, and a fusion adjustment and smoothing module.

[0042] The signal acquisition and alignment module acquires multi-source feedback information and performs unified time stamping during the control of the multi-joint linkage robotic arm. It then performs alignment processing with the same sampling rhythm through the unified time stamp to form an aligned time sequence as the input sequence for subsequent fluctuation comparison.

[0043] The feature extraction and expansion module compares the signal amplitude and phase changes in the input sequence, extracts short-time resonance segments between adjacent time points, generates mixing feature trajectories, and writes the mixing feature trajectories to the end of the input sequence to form an expanded sequence containing mixing features.

[0044] The jump detection module detects the continuous update trend of joint angles along the mixing feature trajectory in the extended sequence, and locates the angle jump segment by combining the tail inertial change limit of the mixing feature trajectory, and generates a jump suspect list containing the jump time and position.

[0045] The signal comparison and separation module reviews the multi-source signal combination patterns in the extended sequence based on the suspected jump list, compares the jump time position with the real motion change trend in the extended sequence, separates pseudo jump segments, and generates a draft interference window.

[0046] The fusion adjustment and smoothing module dynamically adjusts the fusion rhythm in the extended sequence according to the interference window draft. Within the interference window, it adjusts the participation ratio of multi-source feedback information and delays the time when high-risk feedback information enters the fusion rhythm. It introduces local time misalignment to smooth the angle sequence, so that the angle sequence avoids jump input caused by instantaneous mixing.

[0047] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0048] This invention improves motion control stability by performing time alignment, mixing feature extraction, jump segment localization, and interference window adjustment on multi-source feedback information before and after interference generation, thereby maintaining the continuity of joint angle sequences in dynamic environments. Through in-depth analysis of mixing feature trajectories and separation of pseudo-jump segments, the angle update process more closely approximates the actual motion state, enabling multi-joint linkage actions to maintain trajectory smoothness even under high-speed operating conditions, thus improving the accuracy of the end effector position.

[0049] This invention adjusts the participation ratio of multi-source feedback information within the interference window and delays the processing of high-risk feedback information, enabling the fusion rhythm to have a flexible buffering capability during abrupt changes. By using local time misalignment, the angle sequence gradually transitions within the interference window, thereby suppressing the impact of sudden inputs on the entire trajectory following process. This reduces the occurrence of accidental shutdowns during long-term continuous operation of the robotic arm, improving the safety and reliability of coordinated operations. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0051] Figure 1 This is a flowchart of a control method for a high-precision multi-joint linkage robotic arm according to the present invention.

[0052] Figure 2 This is a schematic diagram of the control device for a high-precision multi-joint linkage robotic arm according to the present invention. Detailed Implementation

[0053] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0054] This invention provides, for example Figure 1 The control method for a high-precision multi-joint linkage robotic arm shown includes the following steps:

[0055] During the control of a multi-joint linkage robotic arm, multi-source feedback information is collected and uniformly time-stamped. The uniform time-stamped information is then used to perform alignment processing with the same sampling rhythm to form an aligned time sequence, which serves as the input sequence for subsequent fluctuation comparison.

[0056] In the control of multi-joint robotic arms, precise time synchronization is crucial for efficient control. To ensure that the various joints of the robotic arm can move in a coordinated manner and accurately execute trajectory planning, data from different sensors must be effectively integrated. In particular, the acquisition and alignment of multi-source feedback information can ensure that all signals are analyzed and processed within a unified time frame, thereby avoiding errors and instability caused by signal misalignment. The specific implementation steps are as follows:

[0057] The robotic arm control system collects various feedback information from each joint in real time, including angle encoder signals, acceleration signals, current changes, and torque feedback. This information comes from different sensors, each with different sampling frequencies and transmission delays, making direct comparison and fusion impossible. To address this issue, the system first adds a unified timestamp to the data collected by each sensor. These timestamps not only represent the sampling time of each sensor signal but also reflect the signal's true state at the same moment. Timestamping all sensor data lays the foundation for subsequent data alignment and comparison. This introduction of timestamps enables synchronous comparison and analysis of various signals at the same time, providing a unified input for the robotic arm's control system.

[0058] After unifying the time stamp, the system enters the data alignment stage. Specifically, because different sensors have different sampling frequencies, the time series of the data may also differ, and directly comparing these data could introduce errors. Therefore, the system needs to perform synchronous alignment processing on the acquired time series. This process involves not only unifying the time stamp but also adjusting the sampling rhythm of each sensor signal according to its characteristics, ensuring that the sampling frequencies of all signals are consistent. Through this synchronous processing method, the data from each joint of the robotic arm can be accurately compared within the same time step, allowing subsequent data analysis and processing to be performed based on a unified time step, avoiding inaccurate signal fusion problems caused by time misalignment.

[0059] After data alignment, the system generates an aligned time series containing synchronized data from all sensors, providing a foundation for subsequent fluctuation comparisons. At this point, the changing trends of various feedback signals are reflected on a unified time scale. Next, the system uses this aligned time series as the input sequence for subsequent fluctuation comparisons. By comparing the amplitude and phase changes of these synchronized signals, the system can clearly identify the dynamic characteristics of various signals. The key to this step is that by integrating signals from different sources into the same time frame, the system can more accurately analyze the correlation between signals and potential interference signals, thus providing a basis for further signal processing and control decisions.

[0060] After time alignment and fluctuation comparison processing, the resulting aligned time series can effectively serve subsequent control decisions and motion planning. In practical applications, the motion trajectory of the robotic arm depends not only on the instantaneous feedback of joint angles but also on the continuity and stability of the signal. The aligned time series provides a stable input data foundation for further signal analysis, mixing feature extraction, angle jump detection, and other stages. In this way, the joints of the robotic arm can move collaboratively along a predetermined trajectory, which not only improves the system's response speed but also enhances motion accuracy and stability, avoiding control errors caused by signal misalignment or synchronization problems.

[0061] The amplitude and phase changes of the signal in the input sequence are compared, short-time resonance segments are extracted between adjacent time points, mixing feature trajectories are generated, and the mixing feature trajectories are written to the end of the input sequence to form an extended sequence containing mixing features.

[0062] In the control process of a multi-joint robotic arm, accurately acquiring the dynamic changes of sensor feedback signals and extracting key information is crucial for ensuring the arm's precise movement. Especially when processing complex sensor data, changes in signal amplitude and phase not only reflect the motion state but also reveal potential interference sources. In this context, by comparing the amplitude and phase changes of the signal in the input sequence, the system can accurately extract short-time resonant segments, thereby generating a mixing characteristic trajectory, which is then added to the input sequence to form an extended sequence containing mixing characteristics. The specific implementation steps are as follows:

[0063] After time alignment processing, the acquired multi-source signals form a unified time series. The amplitude and phase changes of these signals reflect the motion state of the robotic arm joints, as well as potential external disturbances or sensor errors. For these signals, at each time point in the time series, the system first compares and analyzes their amplitude changes. Comparing amplitude changes reveals variations in the joint's motion rate and load at different time points, helping to determine if any abnormal motion occurs. Next, comparing phase changes reflects the synchronicity and interrelationships between signals. Inconsistencies in phase changes may indicate signal delays or disturbances, especially in high-speed motion and complex environments where phase changes can be erroneous due to electromagnetic interference, mechanical wear, and other factors. By comparing these changes, the system can accurately analyze the dynamic characteristics of the signals, providing a foundation for subsequent signal processing steps.

[0064] After comparing amplitude and phase changes, the system proceeds to the extraction stage of short-time resonant segments. A resonant segment refers to a time period in the signal where significant fluctuations or oscillations occur; these fluctuations are often related to the rapid movement of the robotic arm or external disturbances. To extract these short-time resonant segments from the input sequence, the system first sets a threshold based on the amplitude fluctuation range of the amplitude and phase changes. When the amplitude or phase change of the signal exceeds this threshold, the system considers the signal change to be a resonant segment. These resonant segments are the parts of the signal that best reflect abrupt changes in motion state, typically related to nonlinear motion of the robotic arm joints or changes in external dynamic loads. The process of extracting short-time resonant segments requires meticulous comparison and analysis between different time points to ensure accurate identification of key fluctuation characteristics in the signal.

[0065] The extracted short-time resonant segments will generate mixing feature trajectories in the time series. These mixing feature trajectories reflect the dynamic characteristics generated by the interaction of multiple signals. These features typically contain multiple frequency components, representing the combined performance of various influencing factors in the signal. The process of generating mixing feature trajectories involves arranging the extracted resonant segments in chronological order and combining their frequency components and amplitude variations to construct a feature sequence describing the signal fluctuation pattern. This feature sequence not only reflects the instantaneous changes of the signal but also demonstrates the coupling relationships of the signal at different frequencies, helping the control system better understand the complex dynamic patterns in the signal. By generating mixing feature trajectories, the system can provide rich dynamic information for subsequent signal processing, interference suppression, and trajectory planning.

[0066] The mixing characteristic trajectory is written to the end of the input sequence, forming an extended sequence containing mixing characteristics. This extended sequence not only contains the dynamic information of the original signal but also incorporates the fluctuation patterns of the mixing characteristics, enabling the subsequent control system to perform more accurate motion prediction and trajectory tracking based on this extended sequence. The mixing characteristic trajectory in the extended sequence provides rich information on signal changes, helping the system identify potential interference signals and make appropriate adjustments during trajectory following. For example, the system can use the mixing characteristic information in the extended sequence to determine whether there are angle jumps caused by signal disturbances or sensor errors, and then take corresponding corrective measures to ensure that the robotic arm can accurately execute the predetermined trajectory. The introduction of the extended sequence provides strong support for subsequent data fusion, motion control, and interference suppression.

[0067] The continuous update trend of joint angles is detected along the mixing feature trajectory in the extended sequence, and the angle jump segment is located by combining the tail inertial change limit of the mixing feature trajectory, and a jump suspect list containing the jump time position is generated.

[0068] In the control of a robotic arm, accurate monitoring of joint angle changes and timely detection of abnormal jumps are crucial for maintaining the continuity and stability of motion. By detecting the continuous update trend of joint angles along the mixing feature trajectory in the extended sequence, and combining this with the inertial change limit of the tail segment of the mixing feature trajectory, possible angle jump segments can be effectively located and identified, thereby generating a list of suspected jumps. This process not only captures anomalies caused by signal interference or sensor errors, but also effectively avoids misjudgments by the control system due to the presence of jump segments, thus ensuring that the robotic arm executes the predetermined trajectory. The specific implementation steps are as follows:

[0069] After processing the extended sequence, the system enters the stage of detecting the continuous update trend of the mixing feature trajectory. At this point, the extended sequence has incorporated the mixing features extracted from the original signal, including key information such as angle and velocity changes. In this stage, the system first performs a continuous analysis of joint angle changes. By analyzing the angle variation trend at each time point, the system can accurately identify normal joint angle update behavior and detect abnormal changes. If the joint angle update rate changes abruptly within certain time periods, or the change trajectory deviates significantly from the normal trend, the system will mark these areas as potential abnormal segments. During the detection process, not only is the change in a single angle value considered, but the dynamic characteristics of joint motion, such as acceleration and velocity changes, are also taken into account to ensure that all possible abnormal jumps can be comprehensively identified.

[0070] The system analyzes the inertial changes at the tail end of the mixing characteristic trajectory along the extended sequence and determines the change limits. The inertial changes at the tail end of the mixing characteristic trajectory typically represent the final stage of joint movement or abnormal fluctuations caused by external load changes, motion resistance, or other factors. At this stage, the system can identify the normal range and amplitude of angle updates by analyzing the inertial changes. Once the angle update trend exceeds the predetermined inertial change limit, it can be determined that a sudden jump may have occurred within that time period. To achieve this, the system establishes a dynamic threshold based on inertial changes to determine which abrupt changes exceed the normal range, thereby locating potential jump segments. Setting the inertial change limit ensures that the system can identify abnormal jumps while avoiding misjudging short-term fluctuations caused by normal external changes.

[0071] Upon detecting potential anomalous jumps, the system will conduct further time-series analysis to precisely pinpoint the temporal location of these jump segments. Specifically, the system will refine the determination of the jump's temporal location based on the anomalous signal points detected in the previous step. By comparing angular data from surrounding time points, the system can analyze the precise start and end times of the jump segment, thus avoiding misinterpreting abrupt changes as normal joint movements. Each suspected jump point is marked and recorded in a suspected jump list. This list not only includes the temporal location of the jump but can also include additional information as needed, such as the joint number where the jump segment is located and possible sources of interference, providing a basis for subsequent analysis and correction.

[0072] Based on the generated list of suspected transitions, further analysis and separation of transition segments are performed. By comparing with other joint data in the extended sequence, the system can determine which transition segments are genuine angular anomalies and which are pseudo-transitions caused by non-motion factors such as signal interference and sensor errors. To achieve this, the system reviews the entire time series, combining data from several preceding and following time points, to identify whether these abrupt changes are synchronized with other signal sources. Based on this comparative analysis, the system can accurately separate pseudo-transition segments and remove them from the list of suspected transitions. Ultimately, the generated list of suspected transitions will accurately record all genuine transition segments and provide clear guidance for subsequent interference suppression and signal correction.

[0073] By reviewing the multi-source signal combination patterns in the extended sequence based on the suspected jump list, comparing the jump time position with the actual motion change trend in the extended sequence, separating pseudo-jump segments, and generating a draft interference window;

[0074] To improve the stability of joint angle sequences, it is necessary to review the time positions in the suspected jump list and, in conjunction with the multi-source signal combination patterns in the extended sequence, conduct an in-depth analysis of the actual motion change trend. This allows for the separation of pseudo-jump segments and the generation of a draft interference window. This process, through refined time-segment review and signal pattern comparison, enhances the control process's anti-interference capability in dynamic environments. The specific implementation steps are as follows:

[0075] After the jump time list is generated, the system reads the jump time positions item by item from the list and reviews the multi-source signal combination patterns in the extended sequence around each jump time position. The extended sequence contains angle change information obtained from previous processing, mixing characteristic trajectories, and the combination trends of the original multi-source signals. This information has good continuity on the time axis. In this stage, the system selects several time points before and after the jump time position to reproduce the combination patterns of all signals within that time period, allowing the dynamic trend of each type of signal to be analyzed one by one. Through this review method, it is possible to observe whether the angle update changes are consistent with the changes of other signals and to capture the interaction relationships between signals. For example, if the angle change corresponding to the jump time position does not show a corresponding trend in acceleration changes, current changes, and torque changes, it can be preliminarily judged that this jump segment is inconsistent with the actual motion state. The key point of this step is to bring each jump time position in the jump time list into the extended sequence for segment-by-segment review, so that the signal analysis is not limited to a single signal, but is based on the overall characteristics of multi-source signals for judgment.

[0076] The system compares the jump time locations one by one with the actual motion change trends in the extended sequence to determine whether the jump time locations are related to the actual motion state. In this stage, the system not only compares the angle change trends but also combines the amplitude changes and time spans of various signals to observe whether these changes reflect a continuous and reasonable motion pattern. For example, when a robotic arm moves along a predetermined trajectory, the angle change usually has a certain continuity and is accompanied by specific velocity and acceleration changes. If the angle change corresponding to the jump time location lacks a trend related to the actual motion change in the extended sequence, such as the absence of a continuously increasing or decreasing trajectory, it can be further confirmed as a potential pseudo-jump segment. Through this comparison process, the system can establish a correlation between the jump time locations in the suspected jump list and the actual motion change trends, thereby identifying jump segments that do not conform to the actual motion pattern. This comparison method makes jump identification more comprehensive, not limited to individual angle jumps but comprehensively observing the entire motion trend, thus improving the reliability of jump segment identification.

[0077] After comparing the jump time position with the actual motion change trend, the system separates the jump segments. To achieve separation, the system continues to review the combination patterns of various signals in the extended sequence, analyzing the signal corresponding to the jump time position together with multi-source signals at surrounding time points. If the jump segment exhibits independent or uncorrelated fluctuations in the multi-source signal change trend of the extended sequence, i.e., it does not form a unified motion pattern with other signals, then the jump segment will be separated into a pseudo-jump segment. During the separation process, the system fully utilizes the comparison results of the actual motion change trend obtained in the previous step, and combines them with the local dynamic changes of the mixing characteristic trajectory in the extended sequence to make the final separated pseudo-jump segments more accurate. This step uses the jump time position as the core, performing a detailed analysis of each jump segment, so that the separation result is entirely based on the actual changes of the signals in the extended sequence.

[0078] After separating the pseudo-jump segments, the system generates an interference window draft based on the separation results. The interference window draft includes the start and end times of the pseudo-jump segments, as well as the combination morphology characteristics of the signals in their corresponding extended sequences. The process of generating the interference window draft involves marking the separated pseudo-jump segments on the timeline, enabling the system to accurately identify these high-interference regions during subsequent dynamic adjustment of the fusion rhythm. The interference window draft not only records the time intervals in which the jumps occur but also includes descriptions of the signal combination morphology within these time intervals. This allows for more effective adjustment of the participation ratio of multi-source signals and the entry timing of high-risk signals during subsequent dynamic adjustment, thereby reducing the impact of these pseudo-jump segments on the final angle sequence. By generating the interference window draft, the system can identify and address potential interference factors that may cause angle jumps in advance, making subsequent control strategies more stable and smooth.

[0079] According to the interference window draft, the fusion rhythm in the extended sequence is dynamically adjusted. Within the interference window, the participation ratio of multi-source feedback information is adjusted and the time of high-risk feedback information entering the fusion rhythm is delayed. Local time misalignment is introduced to smooth the angle sequence so that the angle sequence avoids jump input caused by instantaneous mixing.

[0080] Once the draft interference window has clearly defined the time interval where the pseudo-jump segments occur, the dynamic adjustment of the fusion rhythm becomes a crucial step in suppressing jump inputs. By redistributing the participation ratio of multi-source feedback information within the interference window and delaying the entry time of high-risk feedback information, the angle jumps caused by instantaneous mixing can be further weakened. Furthermore, by introducing time misalignment within a local time range, the angle sequence exhibits a smooth change trend within the interference window, which helps to reconstruct a more stable angle sequence, thus protecting the final trajectory planning from the influence of anomalous signals. The specific implementation steps are as follows:

[0081] After the interference window draft is generated, the system performs a fundamental reconstruction of the fusion rhythm for various feedback information in the extended sequence. At this stage, the system determines the boundary between the interference window and the normal signal interval based on the time interval of the pseudo-jump segments recorded in the interference window draft, and remaps the corresponding feedback information in the extended sequence onto a unified time axis within that time interval. In this way, the extended sequence presents a signal segment independent of the normal interval within the interference window, making this segment adjustable and reconfigurable. During this process, each type of feedback information maintains its original temporal correlation, but its weight in the fusion rhythm is no longer fixed; instead, it dynamically adjusts with the time span of the interference window and the changes in interference characteristics. This mapping not only ensures a clear time reference for subsequent adjustments but also provides a basis for adjusting the participation ratio and delaying time processing, enabling the fusion rhythm to be optimized step-by-step according to changes in the interference window.

[0082] Based on the extended sequence after mapping, the system adjusts the participation ratio of multi-source feedback information within the interference window. In this stage, the system utilizes the feature information of pseudo-jump segments recorded in the interference window draft to identify the contribution trends and stability of various feedback information within the interference window. Accordingly, it reduces the participation ratio of information sources with frequent fluctuations while increasing the participation ratio of information sources that exhibit more stable performance during the interference window stage. For example, in cases where instantaneous mixing causes jumps, angle and current signals often experience amplitude fluctuations; therefore, the system appropriately reduces their participation ratio within the interference window. Other feedback signals, such as internal joint velocity feedback signals, due to their strong inertial characteristics, will obtain a higher participation ratio within the interference window, thus playing a more stable role in the fusion rhythm. Through this dynamic adjustment, the extended sequence can form a more balanced signal combination within the interference window, enabling the system to extract more stable angle change trends from multi-source feedback.

[0083] After adjusting the participation ratio, the system delays the entry time of high-risk feedback information into the fusion rhythm within the interference window. Based on the high-risk feedback information identified in the previous stage, the system redefines the entry time of the sampling points of this type of feedback information on the time axis of the extended sequence, ensuring it doesn't participate in the fusion rhythm at critical moments within the interference window. By delaying the entry time of high-risk feedback information, the system effectively avoids the impact of sudden jumps caused by instantaneous mixing on the continuous changes in the angle sequence. For example, when a sudden increase or decrease in a certain type of feedback information occurs within the interference window, the system appropriately delays the entry time of this change into the fusion rhythm, allowing it to be balanced by more stable information, thereby reducing the disruption to the continuity of the angle sequence. Furthermore, this delay preserves the integrity of the feedback information but changes its timing in the fusion rhythm, creating a natural buffering effect in the time dimension of the entire process.

[0084] By introducing localized temporal misalignment within the interference window, the angle sequence exhibits a smooth state during the fusion rhythm execution. In this stage, the system utilizes the more stable extended sequence foundation formed in the previous sub-steps to subtly misalign various feedback information within the interference window, preventing their time points from completely overlapping and instead creating a natural temporal distribution within a small range. This temporal misalignment allows the angle sequence to exhibit a continuous, gradual change within the interference window, thus avoiding abrupt input jumps caused by instantaneous mixing. For example, when the angle signal at a certain time point is affected by a pseudo-jump, by misaligning and superimposing it with feedback signals from surrounding time points, the final fused angle can present a natural, gradual change process, rather than a hard abrupt change. Ultimately, after this series of adjustments, the angle changes within the extended sequence exhibit a continuous and gentle trend, effectively weakening abrupt inputs within the interference window, thereby ensuring that the robotic arm maintains stable and consistent motion performance during trajectory planning.

[0085] This invention improves motion control stability by performing time alignment, mixing feature extraction, jump segment localization, and interference window adjustment on multi-source feedback information before and after interference generation, thereby maintaining the continuity of joint angle sequences in dynamic environments. Through in-depth analysis of mixing feature trajectories and separation of pseudo-jump segments, the angle update process more closely approximates the actual motion state, enabling multi-joint linkage actions to maintain trajectory smoothness even under high-speed operating conditions, thus improving the accuracy of the end effector position.

[0086] This invention adjusts the participation ratio of multi-source feedback information within the interference window and delays the processing of high-risk feedback information, enabling the fusion rhythm to have a flexible buffering capability during abrupt changes. By using local time misalignment, the angle sequence gradually transitions within the interference window, thereby suppressing the impact of sudden inputs on the entire trajectory following process. This reduces the occurrence of accidental shutdowns during long-term continuous operation of the robotic arm, improving the safety and reliability of coordinated operations.

[0087] This invention provides, for example Figure 2 The control device for a high-precision multi-joint linkage robotic arm shown includes a signal acquisition and alignment module, a feature extraction and expansion module, a jump detection module, a signal comparison and separation module, and a fusion adjustment and smoothing module.

[0088] The signal acquisition and alignment module acquires multi-source feedback information and performs unified time stamping during the control of the multi-joint linkage robotic arm. It then performs alignment processing with the same sampling rhythm through the unified time stamp to form an aligned time sequence as the input sequence for subsequent fluctuation comparison.

[0089] The feature extraction and expansion module compares the signal amplitude and phase changes in the input sequence, extracts short-time resonance segments between adjacent time points, generates mixing feature trajectories, and writes the mixing feature trajectories to the end of the input sequence to form an expanded sequence containing mixing features.

[0090] The jump detection module detects the continuous update trend of joint angles along the mixing feature trajectory in the extended sequence, and locates the angle jump segment by combining the tail inertial change limit of the mixing feature trajectory, and generates a jump suspect list containing the jump time and position.

[0091] The signal comparison and separation module reviews the multi-source signal combination patterns in the extended sequence based on the suspected jump list, compares the jump time position with the real motion change trend in the extended sequence, separates pseudo jump segments, and generates a draft interference window.

[0092] The fusion adjustment and smoothing module dynamically adjusts the fusion rhythm in the extended sequence according to the interference window draft. Within the interference window, it adjusts the participation ratio of multi-source feedback information and delays the time when high-risk feedback information enters the fusion rhythm. It introduces local time misalignment to smooth the angle sequence, so that the angle sequence avoids jump input caused by instantaneous mixing.

[0093] The present invention provides a control method for a high-precision multi-joint linkage robotic arm, which is implemented by the control device for the high-precision multi-joint linkage robotic arm described above. For details of the specific method and process of the control device for the high-precision multi-joint linkage robotic arm, please refer to the embodiment of the control method for the high-precision multi-joint linkage robotic arm described above, which will not be repeated here.

[0094] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A control method for a high-precision multi-joint linkage robotic arm, characterized in that, Includes the following steps: During the control of a multi-joint linkage robotic arm, multi-source feedback information is collected and uniformly time-stamped. The uniform time-stamped information is then used to perform alignment processing with the same sampling rhythm to form an aligned time sequence, which serves as the input sequence for subsequent fluctuation comparison. The amplitude and phase changes of the signal in the input sequence are compared, short-time resonance segments are extracted between adjacent time points, mixing feature trajectories are generated, and the mixing feature trajectories are written to the end of the input sequence to form an extended sequence. The continuous update trend of joint angles is detected along the mixing feature trajectory in the extended sequence, and the angle jump segment is located by combining the tail inertial change limit of the mixing feature trajectory, and a list of suspected jump segments is generated. By reviewing the multi-source signal combination patterns in the extended sequence based on the suspected jump list, comparing the jump time position with the actual motion change trend in the extended sequence, separating pseudo-jump segments, and generating a draft interference window; Based on the interference window draft, the fusion rhythm in the extended sequence is dynamically adjusted. Within the interference window, the participation ratio of multi-source feedback information is adjusted and the time when high-risk feedback information enters the fusion rhythm is delayed. Local time misalignment smoothing angle sequence is introduced.

2. The control method for a high-precision multi-joint linkage robotic arm according to claim 1, characterized in that, The steps for forming the input sequence are as follows: During the control of a multi-joint linkage robotic arm, multi-source feedback information is collected and uniformly time-marked, forming a data set with consistent time points from the uniform time-marked data. In a dataset with consistent time points, the sampling rhythm of different sensors is adjusted, and data with different sampling rhythms are integrated into a time series with a unified sampling rhythm. In a time series with a uniform sampling rhythm, an aligned time series containing multiple types of feedback information is generated. The amplitude and phase changes in the aligned time series are compared and analyzed with time points as references. An input sequence for fluctuation comparison is formed in the aligned time series of the comparative analysis. The input sequence is used as the basis for continuous joint angle updates and participates in the subsequent signal analysis process.

3. The control method for a high-precision multi-joint linkage robotic arm according to claim 2, characterized in that, The steps for forming the extended sequence are as follows: The signal amplitude changes at each time point in the input sequence are compared, and the signal phase changes at the same time point are compared accordingly to form a joint relationship between amplitude changes and phase changes. In the joint change relationship, identify the time interval where amplitude change and phase change fluctuate simultaneously between adjacent time points, and extract short-time resonance segments within this time interval; The short-time resonance segments are arranged in chronological order to form a mixing characteristic trajectory, so that the mixing characteristic trajectory reflects the dynamic change characteristics generated by the interaction of multiple source signals. The mixing feature trajectory is written to the end of the input sequence to form an extended sequence containing the mixing feature trajectory, which is then used in subsequent joint angle change analysis.

4. The control method for a high-precision multi-joint linkage robotic arm according to claim 3, characterized in that, The extraction of short-time resonance segments is based on the time interval in which amplitude and phase changes in the input sequence exhibit synchronous fluctuations at adjacent time points. The mixing feature trajectory records the joint change process of multi-source signals within this time interval and serves as the time correlation information in the extended sequence to reflect the abnormal fluctuation characteristics.

5. The control method for a high-precision multi-joint linkage robotic arm according to claim 3, characterized in that, The steps for generating the suspected jump list are as follows: The trend of joint angle change is continuously updated by following the mixing feature trajectory in the extended sequence, thus identifying the trajectory of joint angle change in the time series. In the continuously updated trend detection results, the inertial change of the tail segment of the mixing feature trajectory is analyzed, and the inertial change limit corresponding to the angle change is determined. Under the constraint of inertial change limit, identify the time interval of angle change exceeding the inertial change limit and locate the corresponding angle jump segment; The time positions corresponding to the angle jump segments are recorded as jump time positions and compiled into a list of suspected jumps for use in subsequent sequence analysis.

6. The control method for a high-precision multi-joint linkage robotic arm according to claim 5, characterized in that, The jump time positions in the suspected jump list are formed by marking the start and end time points of the angular jump segments in the extended sequence and arranging them in chronological order.

7. The control method for a high-precision multi-joint linkage robotic arm according to claim 5, characterized in that, The steps for generating the interference window draft are as follows: By reviewing the combination patterns of multi-source signals in the extended sequence around the jump time positions in the jump suspect list, the signal change trend within the corresponding time interval can be reproduced. In the multi-source signal combination pattern obtained by review, the jump time position is compared with the actual motion change trend in the extended sequence to identify the angle change segment that is inconsistent with the motion change trend. In the comparison results of real motion change trends, inconsistent angle change segments are separated to form a set of pseudo-jump segments. An interference window draft is generated based on the distribution of the pseudo-jump fragment set on the time axis, and the morphological characteristics of the multi-source signal combination within the corresponding time interval are recorded.

8. The control method for a high-precision multi-joint linkage robotic arm according to claim 7, characterized in that, The draft interference window divides the time axis according to the continuous distribution of pseudo-jump segments, and associates the corresponding multi-source signal combination morphological characteristics within each interference window to distinguish between the jump concentration interval and the normal change interval, and to limit the time range in the extended sequence that needs to participate in the fusion rhythm adjustment.

9. The control method for a high-precision multi-joint linkage robotic arm according to claim 7, characterized in that, The steps for dynamically adjusting the fusion rhythm are as follows: Based on the draft interference window, the time interval corresponding to the interference window is determined in the extended sequence, and the multi-source feedback information is processed by time axis mapping within the time interval; In the extended sequence of time-axis mapping processing, the participation ratio of multi-source feedback information is adjusted to form a signal combination form that conforms to the characteristics of the interference window; In the signal combination pattern involved in the ratio adjustment, the time when high-risk feedback information enters the fusion rhythm is delayed, thereby changing the time position of high-risk feedback information in the extended sequence. Local time misalignment is introduced into the extended sequence of delayed processing to smooth angle changes and form an angle sequence that avoids instantaneous mixing jump input.

10. A control device for a high-precision multi-joint linkage robotic arm, used to implement the control method for a high-precision multi-joint linkage robotic arm as described in any one of claims 1-9, characterized in that, It includes a signal acquisition and alignment module, a feature extraction and expansion module, a transition detection module, a signal comparison and separation module, and a fusion adjustment and smoothing module. The signal acquisition and alignment module acquires multi-source feedback information and performs unified time stamping during the control of the multi-joint linkage robotic arm. It then performs alignment processing with the same sampling rhythm through the unified time stamp to form an aligned time sequence as the input sequence for subsequent fluctuation comparison. The feature extraction and expansion module compares the signal amplitude and phase changes in the input sequence, extracts short-time resonance segments between adjacent time points, generates mixing feature trajectories, and writes the mixing feature trajectories to the end of the input sequence to form an expanded sequence. The jump detection module detects the continuous update trend of joint angles along the mixing feature trajectory in the extended sequence, and combines the tail inertial change limit of the mixing feature trajectory to locate the angle jump segment and generate a jump suspect list. The signal comparison and separation module reviews the multi-source signal combination patterns in the extended sequence based on the suspected jump list, compares the jump time position with the real motion change trend in the extended sequence, separates pseudo jump segments, and generates a draft interference window. The fusion adjustment and smoothing module dynamically adjusts the fusion rhythm in the extended sequence based on the interference window draft. Within the interference window, it adjusts the participation ratio of multi-source feedback information and delays the time when high-risk feedback information enters the fusion rhythm, introducing a local time misalignment to smooth the angle sequence.

Images