Multi-axis motor servo control method and system based on distributed feedback

Abstract

The invention provides a multi-shaft motor servo control method and system based on distributed feedback, and relates to the technical field of multi-shaft motor control, and the method comprises the steps: determining an initialization master-slave mode of a plurality of servo shafts of a multi-shaft motor, and monitoring and evaluating the performance evaluation index of each servo shaft in real time; when the performance evaluation index of the leader axis is reduced to a first preset threshold value, triggering switching early warning, and selecting an optimal shift connection axis based on a predetermined competency scoring mechanism; and starting a smooth transition program, transferring the control right of the leader from the current leader shaft to a shift connection shaft in a preset overlapping control period, and adaptively adjusting control parameters based on the new role relationship. According to the method and the device, the technical problem that the control precision of the multi-axis motor is further influenced because the overall stability is deteriorated along with the reduction of the reference performance due to the performance degradation or fault of a single reference node in the prior art can be solved, and the control precision of the multi-axis motor is improved through the dynamic lossless switching of the multi-axis cooperative reference.

Description

Technical Field

[0001] This application relates to the field of multi-axis motor control technology, specifically to a multi-axis motor servo control method and system based on distributed feedback. Background Technology

[0002] In the field of multi-axis motor coordinated control, a pre-defined reference axis is typically used to generate synchronization commands, and the other axes follow these commands to achieve coordinated motion. However, this reference axis can experience performance degradation over long-term operation due to factors such as mechanical wear, temperature rise, or load changes. The decline in key indicators such as tracking error and response speed directly degrades the quality of the issued synchronization commands. Since the motion accuracy of the entire system is tightly coupled to this single reference source, the overall control performance inevitably decreases, forming an inherent bottleneck in system stability and accuracy.

[0003] Currently, solutions to these problems mainly rely on periodic maintenance or reactive shutdowns for replacement after a failure, lacking the ability to monitor and proactively manage the performance of the reference axis in real time during operation. Existing methods cannot provide early warnings in the early stages of reference axis performance degradation, nor can they seamlessly transfer reference functions to other axes with better performance without interrupting the process. This makes it difficult to guarantee the long-term accuracy of the system, and downtime for maintenance seriously affects the efficiency and reliability of continuous production.

[0004] In summary, existing technologies suffer from the technical problem that the overall stability deteriorates as the reference performance decreases due to the performance degradation or failure of a single reference node, which further affects the control accuracy of multi-axis motors. Summary of the Invention

[0005] The purpose of this application is to provide a multi-axis motor servo control method and system based on distributed feedback, in order to solve the technical problem in the prior art where the overall stability deteriorates as the reference performance decreases due to the performance degradation or failure of a single reference node, which further affects the control accuracy of the multi-axis motor.

[0006] To achieve the above objectives, this application provides a multi-axis motor servo control method and system based on distributed feedback.

[0007] Firstly, this application provides a multi-axis motor servo control method based on distributed feedback. This method is implemented through a multi-axis motor servo control system based on distributed feedback. The method includes: determining an initial master-slave mode for multiple servo axes of a multi-axis motor; monitoring and evaluating the performance evaluation index of each servo axis in real time when the multi-axis motor operates according to the initial master-slave mode; triggering a switching warning when the performance evaluation index of the leader axis drops to a first preset threshold; evaluating and selecting the optimal successor axis from all follower axes based on a predetermined competence scoring mechanism; initiating a smooth transition procedure; transferring the leader's control from the current leader axis to the successor axis within a preset overlapping control period; wherein the total control output is a weighted mixture of the current leader axis output and the successor axis output; after the transition is completed, the successor axis is confirmed as the new leader axis, and the control parameters are adaptively adjusted based on the new role relationship.

[0008] Optionally, after power-on, the controllers of multiple servo axes perform local self-tests and broadcast identity declaration messages containing their own hardware identifiers, performance parameters, and initial health status. Based on the identity declaration messages received from other axes, each servo axis conducts initial leader negotiation according to preset election rules to determine a unique leader axis. The preset election rules comprehensively compare static priority configurations, hardware capability indicators, and self-test results. All axes other than the leader axis are automatically configured as follower axes, completing the initialization and generating the initial master-slave mode.

[0009] Optionally, local status data of each servo axis is collected synchronously, including actual position, commanded position, output torque, and winding temperature; dynamic performance sub-indicators of each servo axis are calculated based on the local status data; a weighted fusion function is used to synthesize the dynamic performance sub-indicators into a comprehensive performance evaluation index, and a sliding window average filter is performed to generate the performance evaluation index of each servo axis.

[0010] Optionally, the dynamic performance sub-indicators include the absolute value of instantaneous tracking error, the rate of error change, the torque fluctuation variance, and the temperature rise rate.

[0011] Optionally, when the performance evaluation index of the leader axis falls below the first preset threshold for the first time, the leader axis is marked to enter the warning state; a timer is started, and if the performance evaluation index of the leader axis continues to fall below the first preset threshold within the preset warning observation period, a switch warning is officially triggered; otherwise, the warning state is lifted.

[0012] Optionally, performance evaluation metrics for each follower axis are extracted; the communication delay between each follower axis and the leader axis is obtained, and the performance evaluation metrics of each follower axis are normalized and weighted to generate a competency score for each follower axis; and the follower axis with the highest competency score is selected as the successor axis.

[0013] Optionally, if the follower axis with the highest competence score is lower than the preset minimum acceptance standard, a downgrade process is triggered, and the leader is not switched.

[0014] Optionally, upon entering an overlap control cycle, the current leader axis and the successor axis synchronously load the same future time domain motion trajectory parameters. In the first overlap cycle, the weight coefficient of the current leader axis is set to 1. In each subsequent overlap control cycle, the weight coefficient of the current leader axis decreases linearly according to a preset ramp function, while the weight output by the successor axis increases accordingly. In the last overlap cycle, the weight coefficient of the current leader axis is 0.

[0015] Optionally, the fault-tolerant handling steps for communication anomalies are as follows: During the real-time monitoring and evaluation of the performance evaluation indicators of each servo axis, the communication status of each servo axis is monitored synchronously. If any servo axis loses communication, the corresponding servo axis is marked as an offline axis. If the offline axis is the leader axis, an emergency switchover process is triggered. Based on the most recent performance evaluation indicator data, a successor axis is selected, and a forced switchover is completed using a shortened overlap period. If the offline axis is the follower axis, the leader axis redistributes the load.

[0016] Secondly, this application also provides a multi-axis motor servo control system based on distributed feedback, used to execute the multi-axis motor servo control method based on distributed feedback as described in the first aspect. The multi-axis motor servo control system based on distributed feedback includes: an automatic switching mechanism module, used to determine the initial master-slave mode of multiple servo axes of the multi-axis motor, and to monitor and evaluate the performance evaluation index of each servo axis in real time when the multi-axis motor is running in the initial master-slave mode; a competency scoring module, used to trigger a switching warning when the performance evaluation index of the leader axis drops to a first preset threshold, and to evaluate and select the optimal successor axis from all follower axes based on a predetermined competency scoring mechanism; and a smooth transition module, used to initiate a smooth transition procedure, transferring the leader control from the current leader axis to the successor axis within a preset overlapping control period, wherein the total control output is a weighted mixture of the current leader axis output and the successor axis output. After the transition is completed, the successor axis is confirmed as the new leader axis, and the control parameters are adaptively adjusted based on the new role relationship.

[0017] One or more technical solutions provided in this application have at least the following technical effects or advantages: By determining the initial master-slave mode of multiple servo axes of a multi-axis motor, the performance evaluation index of each servo axis is monitored and evaluated in real time when the multi-axis motor runs in the initial master-slave mode. When the performance evaluation index of the leader axis drops to a first preset threshold, a switching warning is triggered. Based on a predetermined competency scoring mechanism, the optimal successor axis is evaluated and selected from all follower axes. A smooth transition procedure is initiated, and within a preset overlapping control cycle, the leader control is transferred from the current leader axis to the successor axis. The total control output is a weighted mixture of the current leader axis output and the successor axis output. After the transition is completed, the successor axis is confirmed as the new leader axis, and the control parameters are adaptively adjusted based on the new role relationship. In other words, by determining the initial master-slave mode of multiple servo axes of a multi-axis motor, including one leader axis and several follower axes, the performance evaluation indicators of each servo axis are monitored and evaluated in real time. When the performance of the leader axis drops to a certain level, the switchover warning is automatically identified and triggered. The optimal successor axis is evaluated and selected from multiple follower axes for a smooth transition. After the transition is completed, the new leader axis can adaptively adjust the control parameters based on the new role relationship, which improves the servo control accuracy of the multi-axis motor and further enhances the operational reliability and stability of the multi-axis motor.

[0018] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the multi-axis motor servo control method based on distributed feedback proposed in this application.

[0021] Figure 2 This is a schematic diagram of the structure of the multi-axis motor servo control system based on distributed feedback in this application.

[0022] Figure labeling: Automatic switching mechanism module 11, competency scoring module 12, smooth transition module 13. Detailed Implementation

[0023] This application provides a multi-axis motor servo control method and system based on distributed feedback, solving the technical problem in existing technologies where the overall stability deteriorates due to the performance degradation or failure of a single reference node, further affecting the control accuracy of the multi-axis motor. By determining the initial master-slave mode of multiple servo axes of the multi-axis motor, including one leader axis and several follower axes, the performance evaluation indicators of each servo axis are monitored and evaluated in real time. When the performance of the leader axis deteriorates to a certain level, a switching warning is automatically triggered. The optimal successor axis is evaluated and selected from multiple follower axes for a smooth transition. After the transition is completed, the new leader axis can adaptively adjust its control parameters based on the new role relationship, improving the servo control accuracy of the multi-axis motor and further enhancing the operational reliability and stability of the multi-axis motor.

[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0025] Example 1, please refer to the appendix. Figure 1 This application provides a multi-axis motor servo control method based on distributed feedback, wherein the multi-axis motor servo control method based on distributed feedback is applied to a multi-axis motor servo control system based on distributed feedback, and the multi-axis motor servo control method based on distributed feedback specifically includes the following steps: Determine the initial master-slave mode for multiple servo axes of a multi-axis motor. When the multi-axis motor is running in the initial master-slave mode, monitor and evaluate the performance indicators of each servo axis in real time.

[0026] Furthermore, this application also includes the following steps: After power-on, the controllers of multiple servo axes perform local self-tests and broadcast identity declaration messages containing their own hardware identifiers, performance parameters, and initial health status; each servo axis, based on the identity declaration messages received from other axes, conducts initial leader negotiation according to preset election rules to determine a unique leader axis, wherein the preset election rules comprehensively compare static priority configuration, hardware capability indicators, and self-test results; and automatically configures the axes other than the leader axis as follower axes, completes initialization, and generates the initial master-slave mode.

[0027] Specifically, after the system is powered on, it does not immediately begin movement but enters a distributed initialization phase. Each servo axis controller performs a local self-test, automatically diagnosing itself and its connected hardware, such as motors, encoders, and power circuits, checking for short circuits, open circuits, sensor failures, and other faults to ensure there are no major hardware malfunctions. All servo axes simultaneously broadcast their identity declaration messages to the network, essentially acting as electronic IDs for each axis. These messages contain a unique ID, rated power, peak torque, encoder resolution, and other hardware identifiers and performance parameters, as well as the initial health status obtained after the self-test, such as normal or encoder warning. For example, suppose a multi-axis motor system consists of four servo axes: A, B, C, and D. After power-on, each axis performs a self-test and broadcasts an identity declaration message. For instance, axis A's message might include hardware identifier A-001, maximum torque 50 Nm, operating temperature 25°C, and no faults.

[0028] After each servo axis sends its own identity declaration message, it receives identity declaration messages from other axes, thus gaining access to information about all servo axes in the system. Based on the same preset election rules, initial leader negotiation is conducted among the multiple servo axes to determine a unique leader axis. The preset election rules define the criteria for selecting the leader axis and determining the roles of each axis during initialization, using multiple factors (such as hardware priority and self-test results) to determine which servo axis is most suitable to be the leader. Static priority configuration consists of pre-set weight values ​​that do not change during operation, such as specifying that a particular axis has priority to become the leader under equal conditions. Hardware capability indicators are quantitative data reflecting the physical performance of the axis extracted from the identity declaration messages, such as maximum speed, rated torque, and communication bandwidth. Self-test results are qualitative or quantitative conclusions derived from local self-tests, directly assessing whether the axis is currently functioning normally. For example, the preset election rules may include: a completely normal self-test result is required for eligibility; if all are normal, static priorities are compared, with the higher priority winning; if priorities are the same, key hardware capability indicators, such as rated torque, are further compared, with the stronger indicator winning.

[0029] Through a preset election rule, all axes independently arrive at the same conclusion, determining a unique leader axis. Once the leader axis is determined, other servo axes are automatically configured as follower axes, completing initialization and generating an initial master-slave mode. The initial master-slave mode includes one master axis and multiple slave axes. The leader axis generates a global reference trajectory and distributes it to each follower axis. All servo axes perform closed-loop control based on local feedback and distributed feedback data from other axes, and periodically exchange health status data to achieve system state consensus. The leader axis plans and executes its own position / velocity trajectory; follower axes do not plan independently, and their goals are entirely or partially derived from the leader. For example, the follower's target position = the leader's actual position + a fixed offset.

[0030] In master-slave mode, all servo axes periodically exchange their health status data, including temperature, load, and fault status. This data is shared through a distributed feedback mechanism to achieve system state consensus, ensuring that each servo axis can self-adjust based on the global state. This helps identify potential problems, provides early fault warnings, and prevents the entire system from going out of control due to anomalies in a single axis. By selecting a leader axis and establishing master-slave mode, axes can be precisely synchronized, reducing error accumulation and improving the control accuracy of the multi-axis motor system. Each axis can move strictly according to the reference trajectory, avoiding overall control instability caused by deviations in individual axes.

[0031] Furthermore, this application also includes the following steps: synchronously collecting local status data of each servo axis, the local status data including actual position, commanded position, output torque, and winding temperature; calculating dynamic performance sub-indicators for each servo axis based on the local status data; using a weighted fusion function to synthesize the dynamic performance sub-indicators into a comprehensive performance evaluation index, and performing sliding window averaging filtering to generate the performance evaluation index for each servo axis.

[0032] Furthermore, this application also includes the following steps: the dynamic performance sub-indicators include the absolute value of instantaneous tracking error, the error change rate, the torque fluctuation variance, and the temperature rise rate.

[0033] Specifically, after normal operation in the initial master-slave mode, the performance testing and evaluation process is initiated. Local status data for each servo axis is collected synchronously, including actual position, commanded position, output torque, and winding temperature. Actual position is the axis's current position calculated in real-time from the motor encoder, typically expressed in pulse counts, degrees, or millimeters. Commanded position is the position setpoint issued to the axis by the controller in the current cycle, representing the target position the axis is expected to reach. Output torque is the torque value corresponding to the current actually output to the motor windings by the servo driver based on the control algorithm, reflecting the motor's output power. Winding temperature is the real-time temperature value measured by a temperature sensor installed inside the motor stator windings, a key parameter for assessing the motor's thermal load and health status. For example, the commanded position of the X-axis is 150.125mm, the actual position is 150.103mm, the output torque is 5.6Nm, and the winding temperature is 72.3℃; the commanded position of the Y-axis is 85.042mm, the actual position is 85.051mm, the output torque is 4.1Nm, and the winding temperature is 68.7℃; the commanded position of the C-axis is 45.18°, the actual position is 45.15°, the output torque is 2.8Nm, and the winding temperature is 75.1℃.

[0034] The dynamic performance sub-indicators for each servo axis are calculated using local state data, including the absolute value of instantaneous tracking error, error change rate, torque fluctuation variance, and temperature rise rate. The absolute value of instantaneous tracking error is the absolute value of the difference between the commanded position and the actual position in the current control cycle, quantifying the axis's instantaneous tracking accuracy to the command; the unit is the same as position. The error change rate is the difference between the current tracking error and the tracking error of the previous control cycle, divided by the control cycle time. It reflects the trend and speed at which the error accelerates or decreases, used to determine whether the system is stabilizing or diverging. The torque fluctuation variance is a mathematical measure of the dispersion of the output torque value around its average value over the most recent statistical period; a larger variance indicates more unstable torque output, potentially indicating jamming in the mechanical transmission or sudden load changes. The temperature rise rate is the change in winding temperature per unit time; calculating the average rate of temperature rise over the past 10 seconds is more effective than calculating absolute temperature in predicting overheating risks. The absolute value of the instantaneous tracking error is obtained by subtracting the actual position from the commanded position and taking the absolute value; the error change rate is obtained by subtracting the current error from the previous error and dividing by the cycle time; the torque value over a continuous period of time is statistically analyzed, and its variance is calculated to measure the degree of fluctuation; the temperature value over a continuous period of time is linearly fitted to calculate the temperature rise rate.

[0035] A weighted fusion function is used to synthesize the various dynamic performance sub-indicators to obtain a comprehensive performance evaluation index. The instantaneous tracking error absolute value, error change rate, torque fluctuation variance, and temperature rise rate are normalized and mapped to the same dimension interval of 0 to 1, where 0 represents optimal and 1 represents worst. Weights are assigned to each index based on its importance to the overall shaft performance, such as tracking error weight 0.4, error change rate weight 0.3, torque fluctuation weight 0.2, and temperature rise rate weight 0.1. These weighted sums are then used to obtain an initial comprehensive score. The weighted fusion function is: Comprehensive Performance Evaluation Score = w1 * Instantaneous Tracking Error Absolute Value + w2 * Error Change Rate + w3 * Torque Fluctuation Variance + w4 * Temperature Rise Rate, where w1, w2, w3, and w4 are the weight values ​​corresponding to the instantaneous tracking error absolute value, error change rate, torque fluctuation variance, and temperature rise rate, respectively, and w1 + w2 + w3 + w4 = 1.

[0036] To remove the impact of short-term fluctuations, a sliding window average filter is applied to the comprehensive performance evaluation index. The sliding window average filter is a data processing method that maintains a data queue (window) of fixed time length, such as 1000 comprehensive score data points from the most recent second. Each time a new comprehensive score is calculated, it is added to the end of the window, and the oldest data point is removed. Then, the arithmetic mean of all data within the window is calculated as the final output performance evaluation index. This effectively filters out accidental, transient noise interference, resulting in a smooth and stable performance trend. By acquiring and evaluating the status data of each servo axis in real time, the performance of each axis is precisely monitored, potential problems are identified in a timely manner, and control strategies are adjusted, ensuring that the system can maintain high-precision coordinated motion over a long period, especially under high load or long-term operation. By calculating and weighting the dynamic performance sub-indicators, the real-time status of each servo axis is reflected, giving the control system stronger dynamic response capabilities. It can adaptively adjust in actual operation and avoid performance degradation.

[0037] When the performance evaluation index of the leader axis drops to the first preset threshold, a switching warning is triggered. Based on the predetermined competence scoring mechanism, the optimal successor axis is evaluated and selected from all follower axes.

[0038] Furthermore, this application also includes the following steps: when the performance evaluation index of the leader axis is lower than the first preset threshold for the first time, the leader axis is marked to enter the warning state; a timer is started, and if the performance evaluation index of the leader axis continues to be lower than the first preset threshold within the preset warning observation period, a switching warning is formally triggered; otherwise, the warning state is lifted.

[0039] Specifically, in real-time monitoring, the performance evaluation metrics of the leader axis after each cycle update are compared with a first preset threshold. The first preset threshold is a pre-set numerical threshold for judging whether the performance of the leader axis has begun to show a significant and noteworthy decline; it serves as a health warning line. For example, the comprehensive performance evaluation metric has a maximum score of 100, with 100 being the best and 0 being the worst. This threshold is set at 85. When the metric falls below 85, it means that the axis's performance has moved from the excellent range to the range requiring attention.

[0040] When the performance evaluation metric of the leader axis is first detected to be below a first preset threshold, the leader axis is marked into an alert state, and a timer is started to record the elapsed time. An alert observation period is then initiated to continue monitoring the leader axis's performance. If, during this period, the leader axis's performance continues to fall below the first preset threshold, a switchover alert is formally triggered, initiating the leader axis replacement process. If performance recovers during this period, the alert state is lifted, and control of the current leader axis is maintained. If, during the alert observation period, the leader axis's performance still has not recovered to the threshold, a switchover alert is formally triggered, proving that the performance decline is persistent rather than accidental, indicating that the current leader axis is no longer adequate and a predetermined switchover process must be initiated. At this point, the leader axis switchover mechanism is activated, selecting a better-performing follower axis as the new leader axis. If, during the observation period, the leader axis's performance recovers to the threshold, the alert state is lifted, and the current leader axis continues to be used, avoiding unnecessary switching.

[0041] For example, suppose the first preset threshold is set to 82 points, the warning observation period is 500ms, and the control cycle is 1ms. The current leader spindle (C-axis) experiences slight bearing wear due to prolonged heavy cutting. At T=0 milliseconds, the C-axis performance index is monitored at 81.8 points, falling below the 82-point threshold for the first time; the C-axis is immediately marked as being in a warning state, and a 500ms countdown timer is started. From T=100 to 200ms, the C-axis index briefly drops to 81.5 points due to machining a hard point on the workpiece, but after the hard point passes, the index recovers to 82.5 points at T=300ms and stabilizes at around 83 points for the remainder of the observation period. When the timer reaches 500ms, the warning state is lifted because the index is now above the threshold, and everything returns to normal. Alternatively, if from T=0 to 500 milliseconds, the C-axis index remains below the threshold, such as 81.8, 81.7, 81.6, 81.5, ..., showing a slow but stable downward trend. When the timer reaches 500 milliseconds, the condition is met, confirming that the performance degradation is persistent, and a switchover warning is officially triggered.

[0042] By initiating early warning and switching mechanisms at the initial stage of leader axis performance degradation, potential problems can be detected in a timely manner, preventing system out of control due to single axis performance issues. Even if the leader axis fails, switching warnings can promptly transfer leadership to other axes, ensuring the stability of the multi-axis collaborative system.

[0043] Furthermore, this application also includes the following steps: extracting the performance evaluation index of each follower axis; obtaining the communication delay between each follower axis and the leader axis, normalizing and weighting it with the performance evaluation index of each follower axis to generate a competency score for each follower axis, and selecting the follower axis with the highest competency score as the successor axis.

[0044] Furthermore, this application also includes the following steps: if the score of the follower axis with the highest competence rating is lower than the preset minimum acceptance standard, a downgrade process is triggered, and the leader is not switched.

[0045] Specifically, the latest performance evaluation metrics of all follower axes are extracted to form a quantitative score representing the current overall operating status and capability of each follower axis. A higher score generally indicates better tracking accuracy, stability, and thermal performance of that axis. Simultaneously, the communication latency between each follower axis and the leader axis is obtained; this is the round-trip time for network transmission of control command data between the leader axis and each follower axis. This measures the timeliness of the follower axis receiving the leader's commands; lower latency means faster synchronization response.

[0046] Performance evaluation metrics and communication delays are normalized, and raw data with different dimensions and ranges are mathematically processed to map to the same numerical range, such as 0 to 1. Weighting coefficients are determined based on the importance of each factor's contribution to competence. The performance evaluation metrics and communication delays of each follower axis are weighted to obtain the competence score for each follower axis. For example, the normalized weighted formula is: Competence Score = α * Self-PEI - β * Maximum Prediction Deviation - γ * Predicted Energy Impact - δ * Communication Delay, where α, β, γ, and δ are weighting coefficients. After normalization, each factor is multiplied by its respective importance weighting coefficient, and the weighted sum is used to obtain the final competence score for each follower axis. All scores are compared, and the axis with the highest score is selected as the proposed successor axis.

[0047] The highest competency score is compared with a preset minimum acceptable standard as a safety check. If the highest score is lower than the preset minimum acceptable standard, it means that at the current moment, the state of all follower axes is not ideal. Forcing a switch might result in changing from a poorly performing leader to a barely competent or even worse leader, with the risks outweighing the benefits. In this case, a degradation process will be triggered, abandoning the switch, maintaining the current operation, and possibly triggering a higher-level alarm, prompting a manual check.

[0048] For example, assuming a four-axis (leader L, followers F1, F2, F3) material handling system, the competence is calculated using the formula described above, with weights set as: α=0.5, β=0.2, γ=0.2, δ=0.1. The preset minimum acceptance standard is 70 points. Follower F1's score is 0.5*88 - 0.2*0.1*100 - 0.2*0.08*100 - 0.1*0.05*100 = 39.9 points; Follower F2's score is 0.5*92 - 0.2*0.05*100 - 0.2*0.02*100 - 0.1*0.03*100 = 44.3 points; Follower F3's score is 0.5*78 - 0.2*0.15*100 - 0.2*0.20*100 - 0.1*0.10*100 = 31.0 points. The highest scorer is F2 (44.3 points). However, 44.3 points is far below the minimum acceptable standard of 70 points. Therefore, if it is determined that there is no qualified successor, no leadership change will be initiated, a demotion process will be triggered, such as reducing the overall operating speed to ensure safety, and an alarm for no suitable successor axis will be reported.

[0049] By comprehensively evaluating the performance of follower axes, the system accurately determines which axis is best suited to serve as the leader, thereby improving system stability and accuracy. This avoids accuracy issues or system instability caused by selecting an unsuitable leader axis. Automatic dynamic adjustments are made based on the overall competence scores of different follower axes, ensuring a smooth and intelligent replacement when the leader axis performance deteriorates or malfunctions, thus enhancing the system's adaptability.

[0050] A smooth transition procedure is initiated, and within a preset overlapping control period, the leader's control is transferred from the current leader axis to the successor axis. The total control output is a weighted mixture of the current leader axis output and the successor axis output. After the transition is completed, the successor axis is confirmed as the new leader axis, and the control parameters are adaptively adjusted based on the new role relationship.

[0051] Furthermore, this application also includes the following steps: entering an overlap control cycle, the current leader axis and the successor axis synchronously load the same future time domain motion trajectory parameters; in the first overlap cycle, the weight coefficient of the current leader axis is set to 1, and in each subsequent overlap control cycle, the weight coefficient of the current leader axis decreases linearly according to a preset ramp function, while the weight output by the successor axis increases accordingly; in the last overlap cycle, the weight coefficient of the current leader axis is 0.

[0052] Specifically, once the succession axis is determined and the switch is to be executed, a smooth transition procedure is initiated. The smooth transition procedure is a pre-designed sequence of software logic and calculations designed to achieve a seamless, stable, and disturbance-free transfer of control from the current leader axis to the succession axis. Its core objective is to avoid any perceptible trajectory jumps, speed changes, or torque shocks at the moment of switchover.

[0053] At the start of this cycle, the current leader axis and the successor axis synchronously load the same future time-domain motion trajectory parameters. This means that both obtain identical motion trajectories, ensuring that when calculating control commands, they are based on the exact same position and velocity targets, thus fundamentally avoiding inherent contradictions caused by inconsistent targets. During the switch between the leader and successor axes, to ensure a smooth transition, an overlapping control cycle phase is designed. During this phase, both the current leader axis and the successor axis participate in control output simultaneously, gradually completing the transfer of control responsibility. The motion trajectory parameters describe the expected motion state of the servo axis in the future, including position, velocity, and acceleration. During the leader axis switch, the successor axis needs to load the same trajectory parameters as the leader axis to ensure that the motion trajectory after the switch does not change abruptly.

[0054] Once the overlapped control cycle officially begins, a mixed output calculation is performed once in each control cycle. In the first overlapped cycle, the weight coefficient of the current leader axis is set to 1, while the weight coefficient of the successor axis is set to 0. This means that the total control output signal is completely equivalent to the output calculated separately by the current leader axis, and the system behavior is exactly the same as before the switchover. The successor axis is in standby mode but does not output anything.

[0055] Starting from the second overlap cycle, the weighting coefficients are adjusted once in each control cycle according to a preset ramp function. Specifically, the weighting coefficient of the current leader axis decreases linearly with a fixed step size, while the weighting coefficient of the successor axis output increases linearly in sync. Therefore, the total control output at every instant is equal to (leader weight × leader output) + (successor weight × successor output), like a fade-in / fade-out process of control transfer, where the dominance of control smoothly transfers from the old leader to the new leader over time. In the last overlap cycle, the weighting coefficient of the current leader axis decreases to 0, while the weighting coefficient of the successor axis increases to 1. At this point, the total control output signal is entirely generated by the successor axis, and the old leader axis completely exits the control output loop. The ramp function is a mathematical function used to smoothly control the decrease and increase of the control weighting coefficients during the transition. Typically, the ramp function is represented by a linear decreasing curve, used to gradually adjust the distribution ratio of the control signal and avoid sudden changes.

[0056] By employing overlapping control cycles and linear weight adjustments, a smooth transition is achieved between the leader and successor axes, avoiding motion instability or oscillations caused by sudden changes in control signals. High coordination and precision are maintained throughout the transition, ensuring a smooth and accurate motion trajectory. The linear decrease and increase of weight coefficients effectively prevent abrupt changes in control signals between the leader and successor axes, reducing oscillations or errors caused by control signal mismatches, enabling the system to smoothly complete the replacement of the leader axis.

[0057] Furthermore, this application also includes the following steps: fault-tolerant handling steps for communication anomalies: during the real-time monitoring and evaluation of the performance evaluation indicators of each servo axis, the communication status of each servo axis is monitored synchronously. If any servo axis loses communication, the corresponding servo axis is marked as an offline axis. If the offline axis is the leader axis, an emergency switchover process is triggered. Based on the most recent performance evaluation indicator data, a successor axis is selected, and a forced switchover is completed using a shortened overlap period. If the offline axis is the follower axis, the leader axis redistributes the load.

[0058] Specifically, to address sudden network failures, fault-tolerant handling procedures were designed for communication anomalies. Built upon routine performance monitoring processes, this forms a comprehensive fault-tolerant system. While the system monitors the performance metrics of each axis in real time, a parallel communication status monitoring thread also runs, rigorously checking whether each axis returns data as expected in each communication cycle. If a data packet communication loss is detected for a particular axis, the system will not wait or repeatedly attempt to retry; instead, it will immediately take explicit action, marking the axis as offline. Logically, this axis is temporarily removed from the collaborative control cluster, its output is ignored, and its state no longer participates in global decision-making.

[0059] Different fault-tolerance strategies are activated based on the different roles of the offline axes. When the offline axis is the leader axis, the most severe failure occurs because the entire system's collaborative baseline disappears instantly. An emergency switchover process is immediately triggered. There is no time for detailed real-time evaluation; instead, a successor axis is quickly selected based on the most recently cached valid performance evaluation data of each follower axis, such as data updated 50 milliseconds ago, combined with their static priorities. Then, a significantly shortened overlap control cycle is initiated, reducing the time the leader and successor axes share control, such as from the normal 50 milliseconds to 5 milliseconds, to quickly complete a forced switchover. This process sacrifices some smoothness for speed, aiming to restore leadership authority in a very short time and prevent the system from spiraling out of control due to a lack of leadership.

[0060] When the offline axis is a follower axis, the system's command core (leader axis) remains intact, but one execution unit is lost. Upon detecting this situation, the leader axis immediately initiates a load redistribution algorithm. For example, in a multi-axis collaborative handling scenario, if an auxiliary support axis goes offline, the leader axis recalculates the motion trajectories and output ratios of the remaining axes to maintain load balance and stability. In a system with multiple motors driving the same load, torque is redistributed, with the remaining healthy axes sharing the original load share of the offline axis. The overall task continues, but in a degraded mode, such as at reduced speed.

[0061] By designing fault-tolerant mechanisms, the system can quickly respond and take appropriate measures when communication failures occur on servo axes, preventing the failure of the entire system due to the failure of a single axis. In particular, when the leader axis fails, the system can quickly switch over, ensuring stable operation.

[0062] In summary, the multi-axis motor servo control method based on distributed feedback provided in this application has the following technical effects: By determining the initial master-slave mode of multiple servo axes of a multi-axis motor, the performance evaluation index of each servo axis is monitored and evaluated in real time when the multi-axis motor runs in the initial master-slave mode. When the performance evaluation index of the leader axis drops to a first preset threshold, a switching warning is triggered. Based on a predetermined competency scoring mechanism, the optimal successor axis is evaluated and selected from all follower axes. A smooth transition procedure is initiated, and within a preset overlapping control cycle, the leader control is transferred from the current leader axis to the successor axis. The total control output is a weighted mixture of the current leader axis output and the successor axis output. After the transition is completed, the successor axis is confirmed as the new leader axis, and the control parameters are adaptively adjusted based on the new role relationship. In other words, by determining the initial master-slave mode of multiple servo axes of a multi-axis motor, including one leader axis and several follower axes, the performance evaluation indicators of each servo axis are monitored and evaluated in real time. When the performance of the leader axis drops to a certain level, the switchover warning is automatically identified and triggered. The optimal successor axis is evaluated and selected from multiple follower axes for a smooth transition. After the transition is completed, the new leader axis can adaptively adjust the control parameters based on the new role relationship, which improves the servo control accuracy of the multi-axis motor and further enhances the operational reliability and stability of the multi-axis motor.

[0063] Example 2: Based on the same inventive concept as the multi-axis motor servo control method based on distributed feedback in Example 1, this application also provides a multi-axis motor servo control system based on distributed feedback. Please refer to the appendix. Figure 2 The multi-axis motor servo control system based on distributed feedback includes: The automatic switching mechanism module 11 is used to determine the initial master-slave mode of multiple servo axes of the multi-axis motor. When the multi-axis motor runs in the initial master-slave mode, it monitors and evaluates the performance evaluation index of each servo axis in real time. The competence scoring module 12 is used to trigger a switching warning when the performance evaluation index of the leader axis drops to a first preset threshold. Based on a predetermined competence scoring mechanism, it evaluates and selects the optimal successor axis from all follower axes. The smooth transition module 13 is used to start a smooth transition program. Within a preset overlapping control cycle, the leader control is transferred from the current leader axis to the successor axis. The total control output is a weighted mixture of the current leader axis output and the successor axis output. After the transition is completed, the successor axis is confirmed as the new leader axis, and the control parameters are adaptively adjusted based on the new role relationship.

[0064] Furthermore, the automatic switching mechanism module 11 in the multi-axis motor servo control system based on distributed feedback is also used for: after power-on, the controllers of multiple servo axes perform local self-tests and broadcast identity declaration messages containing their own hardware identifiers, performance parameters, and initial health status; each servo axis, based on the identity declaration messages received from other axes, conducts initial leader negotiation according to preset election rules to determine a unique leader axis, wherein the preset election rules comprehensively compare static priority configuration, hardware capability indicators, and self-test results; and automatically configures other axes besides the leader axis as follower axes, completes initialization, and generates the initial master-slave mode.

[0065] Furthermore, the automatic switching mechanism module 11 in the multi-axis motor servo control system based on distributed feedback is also used to: synchronously collect local state data of each servo axis, the local state data including actual position, commanded position, output torque, and winding temperature; calculate the dynamic performance sub-indices of each servo axis based on the local state data; use a weighted fusion function to synthesize the dynamic performance sub-indices into a comprehensive performance evaluation index, and perform sliding window averaging filtering to generate the performance evaluation index of each servo axis.

[0066] Furthermore, the automatic switching mechanism module 11 in the multi-axis motor servo control system based on distributed feedback is also used for: the dynamic performance sub-indicators include the absolute value of instantaneous tracking error, error change rate, torque fluctuation variance, and temperature rise rate.

[0067] Furthermore, the competency scoring module 12 in the multi-axis motor servo control system based on distributed feedback is also used to: mark the leader axis as entering an early warning state when the performance evaluation index of the leader axis is lower than the first preset threshold for the first time; start a timer, and if the performance evaluation index of the leader axis continues to be lower than the first preset threshold within the preset early warning observation period, then formally trigger a switching early warning; otherwise, the early warning state is lifted.

[0068] Furthermore, the competency scoring module 12 in the multi-axis motor servo control system based on distributed feedback is also used to: extract the performance evaluation index of each follower axis; obtain the communication delay between each follower axis and the leader axis, normalize and weight it with the performance evaluation index of each follower axis, generate the competency score of each follower axis, and select the follower axis with the highest competency score as the successor axis.

[0069] Furthermore, the competency scoring module 12 in the multi-axis motor servo control system based on distributed feedback is also used to: if the score of the follower axis with the highest competency score is lower than the preset minimum acceptance standard, trigger a downgrade process and do not switch leaders.

[0070] Furthermore, the smooth transition module 13 in the multi-axis motor servo control system based on distributed feedback is also used for: entering an overlapping control cycle, where the current leader axis and the successor axis synchronously load the same future time domain motion trajectory parameters; in the first overlapping cycle, setting the weight coefficient of the current leader axis to 1, and in each subsequent overlapping control cycle, linearly decreasing the weight coefficient of the current leader axis according to a preset ramp function, while the weight output by the successor axis increases accordingly; in the last overlapping cycle, the weight coefficient of the current leader axis is 0.

[0071] Furthermore, the smooth transition module 13 in the multi-axis motor servo control system based on distributed feedback is also used for: fault-tolerant processing steps for communication anomalies: during the real-time monitoring and evaluation of the performance evaluation indicators of each servo axis, the communication status of each servo axis is monitored synchronously. If any servo axis loses communication, the corresponding servo axis is marked as an offline axis. If the offline axis is the leader axis, an emergency switching process is triggered. Based on the most recent performance evaluation indicator data, a successor axis is selected, and a forced switching is completed using a shortened overlap period. If the offline axis is the follower axis, the leader axis redistributes the load.

[0072] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The multi-axis motor servo control method and specific examples based on distributed feedback in the aforementioned embodiment one are also applicable to the multi-axis motor servo control system based on distributed feedback in this embodiment. Through the foregoing detailed description of the multi-axis motor servo control method based on distributed feedback, those skilled in the art can clearly understand the multi-axis motor servo control system based on distributed feedback in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0073] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0074] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A multi-axis motor servo control method based on distributed feedback, characterized in that, include: Determine the initial master-slave mode of multiple servo axes of a multi-axis motor, and monitor and evaluate the performance evaluation indicators of each servo axis in real time when the multi-axis motor is running in the initial master-slave mode. When the performance evaluation index of the leader axis drops to the first preset threshold, a switching warning is triggered. Based on the predetermined competence scoring mechanism, the optimal successor axis is evaluated and selected from all follower axes. A smooth transition procedure is initiated, and within a preset overlapping control period, the leader's control is transferred from the current leader axis to the successor axis. The total control output is a weighted mixture of the current leader axis output and the successor axis output. After the transition is completed, the successor axis is confirmed as the new leader axis, and the control parameters are adaptively adjusted based on the new role relationship.

2. The multi-axis motor servo control method based on distributed feedback as described in claim 1, characterized in that, Determine the initial master-slave mode for multiple servo axes of a multi-axis motor, including: After power-on, the controllers of multiple servo axes perform local self-tests and broadcast identity declaration messages containing their own hardware identifiers, performance parameters, and initial health status. Each servo axis, based on the identity declaration messages received from other axes, conducts initial leader negotiation according to preset election rules to determine a unique leader axis. The preset election rules comprehensively compare static priority configuration, hardware capability indicators, and self-test results. Automatically configure all axes except the leader axis as follower axes to complete initialization and generate the initial master-slave mode.

3. The multi-axis motor servo control method based on distributed feedback as described in claim 1, characterized in that, When the multi-axis motor is running in the initial master-slave mode, the performance evaluation indicators of each servo axis are monitored and evaluated in real time, including: Synchronously collect local status data for each servo axis, including actual position, commanded position, output torque, and winding temperature; Calculate the dynamic performance sub-indicators for each servo axis based on the local status data; A weighted fusion function is used to synthesize the various dynamic performance sub-indicators into a comprehensive performance evaluation index, and a sliding window average filter is applied to generate the performance evaluation index for each servo axis.

4. The multi-axis motor servo control method based on distributed feedback as described in claim 3, characterized in that, The dynamic performance sub-indicators include the absolute value of instantaneous tracking error, error change rate, torque fluctuation variance, and temperature rise rate.

5. The multi-axis motor servo control method based on distributed feedback as described in claim 1, characterized in that, When the performance evaluation index of the leader axis drops to the first preset threshold, a switching warning is triggered, including: When the performance evaluation index of the leader axis falls below the first preset threshold for the first time, the leader axis is marked to enter the warning state. Start a timer. If the performance evaluation index of the leader axis continues to be lower than the first preset threshold within the preset warning observation period, a switchover warning will be officially triggered; otherwise, the warning status will be lifted.

6. The multi-axis motor servo control method based on distributed feedback as described in claim 5, characterized in that, Based on a predetermined competency scoring mechanism, the optimal successor axis is evaluated and selected from all follower axes, including: Extract performance evaluation indicators for each follower axis; The communication delay between each follower axis and the leader axis is obtained, and normalized and weighted with the performance evaluation index of each follower axis to generate a competency score for each follower axis. The follower axis with the highest competency score is selected as the successor axis.

7. The multi-axis motor servo control method based on distributed feedback as described in claim 6, characterized in that, If the follower axis with the highest competence score falls below the preset minimum acceptance standard, a downgrade process is triggered, and the leader is not switched.

8. The multi-axis motor servo control method based on distributed feedback as described in claim 1, characterized in that, Initiate a smooth transition procedure, within a preset overlapping control period, to transfer leadership control from the current leader axis to the succession axis, including: Upon entering the overlapping control cycle, the current leader axis and the successor axis synchronously load the same future time domain motion trajectory parameters. In the first overlapping cycle, the weight coefficient of the current leader axis is set to 1. In each subsequent overlapping control cycle, the weight coefficient of the current leader axis decreases linearly according to the preset ramp function, while the weight output of the successor axis increases accordingly. In the last overlapping period, the weight coefficient of the current leader axis is 0.

9. The multi-axis motor servo control method based on distributed feedback as described in claim 1, characterized in that, It also includes fault-tolerant handling steps for communication anomalies: During the real-time monitoring and evaluation of the performance metrics of each servo axis, the communication status of each servo axis is monitored simultaneously. If any servo axis loses communication, the corresponding servo axis is marked as an offline axis. If the offline axis is the leader axis, an emergency switchover process is triggered. Based on the most recent performance evaluation data, a successor axis is selected, and a shortened overlap period is used to complete the forced switchover. If the offline axis is a follower axis, then the leader axis redistributes the load.

10. A multi-axis motor servo control system based on distributed feedback, characterized in that, The step of implementing the multi-axis motor servo control method based on distributed feedback according to any one of claims 1 to 9, wherein the multi-axis motor servo control system based on distributed feedback comprises: An automatic switching mechanism module is used to determine the initial master-slave mode of multiple servo axes of a multi-axis motor. When the multi-axis motor is running in the initial master-slave mode, the module monitors and evaluates the performance evaluation indicators of each servo axis in real time. The competence scoring module is used to trigger a switching warning when the performance evaluation index of the leader axis drops to the first preset threshold. Based on the predetermined competence scoring mechanism, it evaluates and selects the optimal successor axis from all follower axes. The smooth transition module is used to initiate a smooth transition procedure. Within a preset overlapping control period, the leader's control is transferred from the current leader axis to the successor axis. The total control output is a weighted mixture of the current leader axis output and the successor axis output. After the transition is completed, the successor axis is confirmed as the new leader axis, and the control parameters are adaptively adjusted based on the new role relationship.

Images