A Robot Joint Motor Cooperative Drive Method Based on Dynamic Decoupling of Dual Encoders

Abstract

The application relates to the technical field of robot motor control, in particular to a robot joint motor cooperative driving method based on double-encoder dynamic decoupling. The application effectively overcomes the interference of transmission chain elastic deformation on control precision through real-time comparison and lag compensation mechanism of double-angle signals at the robot motor end and the load end, significantly improves the robot joint trajectory tracking stability under variable load working conditions, can isolate the transmission chain deformation interference in real time under the load working condition of the robot joint main motor, makes the robot joint main motor torque instruction adapt to the load disturbance synchronously, and suppresses the end trajectory deviation outside the influence of unmeasurable external force.

Description

Technical Field

[0001] This invention relates to the field of robot motor control technology, and in particular to a method for collaborative driving of robot joint motors based on dynamic decoupling of dual encoders. Background Technology

[0002] Currently, the joint motor of a robot is the core execution unit of the power system of an industrial robot, and its performance directly determines the positioning accuracy, dynamic response and reliability of the whole machine.

[0003] The existing robot joint motors currently suffer from the following technical problems:

[0004] 1. Currently, traditional robot joint motors rely solely on a single encoder to measure the position of their output shaft. This makes it impossible to perceive the torsional deformation of the reducer and the actual position of the load end in real time. Consequently, the elastic deformation of the transmission chain is completely unmeasurable. Trajectory deviations caused by external force disturbances (such as workpiece collisions / sudden changes in gravity) are passively fed back to the motor control loop, resulting in technical problems such as system response lag and deterioration of positioning accuracy. This makes it impossible for the robot joint motor torque command to synchronously adapt to load disturbances, resulting in trajectory deviations of the robot end effector exceeding ±0.5°.

[0005] 2. Currently, in industrial robots that use hollow integrated joints, the traditional hysteresis compensators embedded in their joint controllers employ fixed delay parameters, which can easily lead to the following problems under varying load / temperature conditions:

[0006] Scenario 1, Low load + low temperature condition: Overcompensation → Excites high frequency resonance (>10Hz) → System oscillation.

[0007] Scenario 2, High load + high temperature conditions: Insufficient compensation → accumulation of elastic deformation energy in the reducer → induction of low frequency resonance (2-8Hz) → fatigue damage to the reducer.

[0008] Scenario 3: Resonant energy cannot be actively canceled → an additional 15%-30% increase in servo energy consumption.

[0009] Therefore, a robot joint motor cooperative drive method based on dynamic decoupling of dual encoders is needed to solve the above problems. Summary of the Invention

[0010] This invention provides a robot joint motor cooperative drive method based on dynamic decoupling of dual encoders. By comparing the dual angle signals at the robot motor end and the load end in real time and using a hysteresis compensation mechanism, this invention effectively overcomes the interference of transmission chain elastic deformation on control accuracy, significantly improves the stability of robot joint trajectory tracking under variable load conditions, and can isolate transmission chain deformation interference in real time under the load condition of the robot joint main motor, so that the torque command of the robot joint main motor can synchronously adapt to the load disturbance, and suppress the end trajectory deviation outside the influence of unmeasurable external forces.

[0011] The technical solution adopted by this invention to solve the above-mentioned technical problems is: a robot joint motor cooperative drive method based on dynamic decoupling of dual encoders, comprising the following steps:

[0012] Step 1: The first angle signal of the robot motor output shaft is acquired in real time by the first encoder installed on the robot motor output shaft;

[0013] Step 2: Acquire the second angle signal in real time using the second encoder installed at the load output end of the robot joint;

[0014] Step 3: Generate a reference torque control command for the robot joint main motor based on the first angle signal and transmit it to the robot joint main motor controller;

[0015] Step 4: Perform hysteresis compensation on the second angle signal to generate a hysteresis compensation signal;

[0016] Step 5: Generate the robot joint auxiliary motor compensation torque control command based on the hysteresis compensation signal and transmit it to the robot joint auxiliary motor controller;

[0017] Step 6: Calculate the difference between the second angle signal and the first angle signal in real time. When the absolute value of this difference exceeds a preset offset threshold:

[0018] If the second angle signal is greater than the first angle signal, the output polarity of the control command for the auxiliary motor compensation torque of the robot joint is negative.

[0019] If the second angle signal is less than the first angle signal, then the output polarity of the control command for the auxiliary motor compensation torque of the robot joint is positive.

[0020] Furthermore, the first angle signal is the real-time rotation angle of the robot motor output shaft, and the second angle signal is the actual position angle of the robot joint end effector.

[0021] Furthermore, the lag compensation operation in step 4 includes:

[0022] Step 4-1: Set a dynamically adjustable delay time window for the robot joint auxiliary motor compensation torque command. The reference value of this time window is 0.8 to 1.5 times the closed-loop control cycle of the robot joint main motor position.

[0023] Step 4-2: Query the preset mapping table to obtain the delay coefficient based on the real-time load rate of the robot joint, and adjust the delay coefficient based on the temperature of the robot joint's main motor;

[0024] Step 4-3: Multiply the delay time window by the delay coefficient and apply it to the second angle signal.

[0025] Furthermore, the logic for setting the delay coefficient in step 4-2 is as follows:

[0026] When the robot joint load rate is between 0% and 30%, the delay factor is set to 0.9;

[0027] When the robot joint load rate is between 30% and 70%, the delay factor is set to 1.0;

[0028] When the robot joint load rate is between 70% and 100%, the delay factor is set to 1.2;

[0029] The delay coefficient is updated at a frequency of 10 to 50 times per second.

[0030] Furthermore, step 4-2, adjusting the delay coefficient based on the temperature of the robot joint's main motor, includes:

[0031] Step 4-2-1: Monitor the temperature of the robot's main joint motor in real time. When the temperature exceeds 65℃ but does not exceed 85℃:

[0032] If the current robot joint load rate is greater than or equal to 50%, the delay coefficient will be increased by an additional 0.1.

[0033] If the current robot joint load rate is less than 50%, the delay factor will be increased by an additional 0.05.

[0034] When the temperature of the main motor of the robot joint exceeds 85°C, the delay coefficient under the load rate of all robot joints is forcibly set to 1.3 times the original value.

[0035] Furthermore, the delay coefficient is a proportional factor used to adjust the length of the delay time applied to the second angle signal of the robot joint during the hysteresis compensation operation.

[0036] Furthermore, the preset offset threshold in step 6 is the actual torsional angle deviation between the robot joint load output end and the robot motor output shaft, and the preset offset threshold range is 0.3° to 0.8°.

[0037] Furthermore, the preset offset threshold setting in step 6 includes: dividing the robot joint load rate range and associating it with the threshold range; when the real-time load rate of the robot joint is less than or equal to 30%, the offset threshold range is set to 0.7° to 0.8°; when the real-time load rate of the robot joint is greater than 30% and less than or equal to 70%, the offset threshold range is set to 0.5° to 0.6°; when the real-time load rate of the robot joint is greater than 70%, the offset threshold range is set to 0.3° to 0.4°.

[0038] Furthermore, in step 3, when generating the robot joint main motor reference torque control command based on the first angle signal, the signal mismatch of the dual encoders is dynamically decoupled through a preset robot joint load disturbance model. This robot joint load disturbance model satisfies the following formula: ;

[0039] Where τ main This provides the reference torque command for the main motor of the robot joint.

[0040] Where θ error The first angle signal is the deviation from the target angle, which is the theoretical angle position that the robot joint motor output shaft should reach at the current moment.

[0041] Where M load This represents the real-time joint load rate of the robot joints.

[0042] Where ω is the stiffness attenuation factor of the robot joint main motor temperature mapping.

[0043] The advantages of this invention are:

[0044] 1. This invention effectively overcomes the interference of transmission chain elastic deformation on control accuracy by using real-time comparison of dual angle signals from the robot motor end and the load end and a hysteresis compensation mechanism. It significantly improves the stability of robot joint trajectory tracking under variable load conditions. Under the load condition of the robot joint main motor, it can isolate transmission chain deformation interference in real time, so that the torque command of the robot joint main motor can synchronously adapt to load disturbances and suppress the end trajectory deviation outside the influence of unmeasurable external forces.

[0045] 2. This invention features a highly adaptive hysteresis compensation mechanism. Addressing the issue of fixed-delay compensators easily inducing resonance under varying loads (operating conditions), this invention dynamically sets the delay coefficient based on real-time joint load rate and adjusts it online in conjunction with the robot joint main motor temperature, constructing a condition-adaptive compensation strength firewall. Under low-load, low-temperature conditions, the compensation strength is reduced to avoid overcompensation and eliminate induced high-frequency resonance. Under high-load, high-temperature conditions, compensation is significantly enhanced to fully offset the accumulation of elastic deformation energy, thereby effectively suppressing low-frequency resonance that induces reducer fatigue damage. The compensation torque generated in real-time by the robot joint auxiliary motor actively offsets deformation energy, converting the 15%-30% additional servo energy consumption caused by resonance accumulation in traditional solutions into effective mechanical work output, significantly reducing the total system energy consumption.

[0046] 3. This invention enables closed-loop protection of the transmission chain status, extending the lifespan of critical components. It not only improves control precision and stability but also proactively constructs a health protection mechanism for the transmission chain (especially the reducer) at the control layer. By using temperature monitoring as a key input for compensation intensity, the compensation intensity is dynamically enhanced when the temperature of the robot joint's main motor abnormally rises, especially with a forced and significant increase at high temperatures, effectively blocking the risk of plastic deformation caused by thermal effects. Through the combined effect of temperature-correlated compensation enhancement and intelligent torque polarity switching, this invention forms a closed-loop protection against transmission chain deformation, significantly improving the fatigue life of critical components such as harmonic reducers. Detailed Implementation

[0047] The technical solution of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Example 1:

[0049] This invention proposes a robot joint motor cooperative drive method based on dynamic decoupling using dual encoders, comprising the following steps: acquiring a first angle signal of the robot motor output shaft in real time using a first encoder installed on the robot motor output shaft; acquiring a second angle signal in real time using a second encoder installed on the robot joint load output end; generating a reference torque control command for the robot joint main motor based on the first angle signal and transmitting it to the robot joint main motor controller; performing hysteresis compensation on the second angle signal to generate a hysteresis compensation signal; generating a compensation torque control command for the robot joint auxiliary motor based on the hysteresis compensation signal and transmitting it to the robot joint auxiliary motor controller; calculating the difference between the second angle signal and the first angle signal in real time, and when the absolute value of the difference exceeds a preset offset threshold: if the second angle signal is greater than the first angle signal, controlling the output polarity of the compensation torque control command for the robot joint auxiliary motor to be negative; if the second angle signal is less than the first angle signal, controlling the output polarity of the compensation torque control command for the robot joint auxiliary motor to be positive.

[0050] The first encoder is a sensor installed on the output shaft of the robot motor to measure the rotation angle. It can be implemented using an incremental photoelectric encoder or an absolute magnetic encoder, and is used to acquire the theoretical output angle of the motor shaft in real time. The second encoder is a position detection device installed at the load output end of the robot joint. It can be implemented using a hollow absolute encoder and is used to directly measure the actual position angle of the end effector. The hysteresis compensation operation refers to the process of applying an adjustable time delay to the load-end angle signal. This can be achieved through a variable delay window implemented using a digital signal processing chip, dynamically eliminating phase hysteresis caused by elastic deformation of the transmission chain. The polarity adjustment of the compensation torque control command refers to the control logic that changes the sign of the torque output based on the direction of the angle difference. This can be achieved through a comparator circuit or software condition judgment, and is used to actively counteract the mechanical energy accumulated by the elastic deformation of the reducer.

[0051] This invention uses real-time dynamic comparison of angle signals from dual encoders at the motor and load ends, combined with an adjustable parameter hysteresis compensation mechanism and active torque polarity control, to form a closed-loop decoupled control architecture for coordinated drive of main and auxiliary motors. While maintaining the reference torque output of the main motor, the auxiliary motor compensates for the trajectory deviation caused by the deformation of the transmission chain in real time, thereby achieving precise position synchronization under load disturbance.

[0052] The working process and principle of this invention are as follows: A first encoder and a second encoder are installed on the output shaft of the robot motor and the joint load output end, respectively, to collect angle signals at the two positions in real time. A reference torque control command for the main motor is generated based on the first angle signal, and a compensation torque control command for the auxiliary motor is generated after hysteresis compensation of the second angle signal. The difference between the two angle signals is calculated in real time, and when the difference exceeds a preset threshold, the output polarity of the auxiliary motor's compensation torque is controlled according to the sign of the difference. This dual-encoder dynamic decoupling method can monitor the elastic deformation of the transmission chain in real time and offset the trajectory deviation caused by deformation through coordinated drive of the main and auxiliary motors. The hysteresis compensation operation ensures that the compensation command and the deformation process of the transmission chain are phase-matched, avoiding compensation misalignment caused by signal transmission delay. By comparing the two angle differences in real time and intelligently switching the polarity of the compensation torque, it is ensured that elastic deformation is actively offset in the early stages of accumulation, blocking the resonant energy transfer path.

[0053] As a preferred embodiment, the solution of the present invention is specifically implemented as follows:

[0054] A high-precision incremental photoelectric encoder is installed on the motor output shaft of the robot joint as the first encoder, and a magnetic encoder is installed on the joint end effector as the second encoder. The first encoder has a resolution of 10,000 lines per revolution, and the second encoder has a resolution of 4,096 lines per revolution.

[0055] The control system acquires angle signals from two encoders at a frequency of 1kHz. Based on the deviation between the first angle signal and the target angle, and in conjunction with preset PID parameters, a reference torque control command for the main motor is generated.

[0056] The second angle signal is compensated for hysteresis, with the compensation time window set to 1.2 times the closed-loop control cycle of the main motor position. The compensated signal is then processed by a bandpass filter to remove high-frequency noise and low-frequency drift.

[0057] Based on the second angle signal after hysteresis compensation, a feedforward-feedback composite control algorithm is used to generate the auxiliary motor compensation torque command. The feedforward part estimates the inertial torque based on angular velocity and angular acceleration, while the feedback part uses fuzzy PID control.

[0058] The difference between the two angle signals is calculated in real time, and the offset threshold is set to 0.5°. When the absolute value of the difference exceeds the threshold, if the second angle signal is greater than the first angle signal, the output polarity of the auxiliary motor compensation torque command is set to negative; otherwise, it is set to positive.

[0059] Control commands for the main and auxiliary motors are transmitted to their respective drivers via a CAN bus. The drivers employ vector control to achieve high-precision torque output.

[0060] Through the above-described scheme, this invention can monitor and compensate for the elastic deformation of the transmission chain in real time, significantly improving the trajectory tracking accuracy of robot joints. The coordinated drive of the main and auxiliary motors enables rapid response during sudden load changes, suppressing system oscillations. The dynamically adjustable hysteresis compensation mechanism matches the compensation command with the actual deformation process, avoiding under- or over-compensation. The intelligently switching compensation torque polarity actively counteracts elastic deformation energy, effectively preventing resonance. This method not only improves the robot's positioning accuracy and dynamic response capability but also extends the service life of key components such as the reducer.

[0061] The present invention further proposes that the first angle signal is the real-time rotation angle of the robot motor output shaft, and the second angle signal is the actual position angle of the robot joint end effector.

[0062] The first angle signal is directly acquired by an encoder mounted on the motor output shaft, reflecting the theoretical motion state of the drive end; the second angle signal is acquired by an encoder at the load output end, characterizing the actual position of the end effector. These two signals correspond to the position information of the drive end and the load end, respectively, forming the basic data source for closed-loop feedback. The measurement accuracy of the actual position angle of the end effector directly affects the effectiveness of generating the compensation torque command.

[0063] Specifically, the motor output shaft encoder collects the rotation angle in real time as the input to the main motor's reference torque control command, while the actual position angle collected by the end effector encoder is compensated for hysteresis to generate the auxiliary motor's compensation torque command. Through real-time comparison of the two angle signals, the mismatch caused by elastic deformation of the transmission chain is dynamically decoupled. For example, when the deviation between the actual position angle of the end effector and the motor output shaft angle exceeds a threshold, the auxiliary motor actively cancels the deformation by switching polarity. This solution suppresses the end effector trajectory deviation to within ±0.3°, improving positioning accuracy by 40% compared to traditional single-encoder solutions.

[0064] As a preferred embodiment, the solution of the present invention is specifically implemented as follows:

[0065] The first angle signal is the real-time rotation angle of the robot motor output shaft, and the second angle signal is the actual position angle of the robot joint end effector. Specifically, the first encoder is installed on the output shaft of the robot joint motor to acquire the rotation angle of the motor output shaft in real time. The second encoder is installed on the end effector of the robot joint to acquire the actual position angle of the end effector in real time. Thus, by using the two encoders in conjunction, the angle information of the robot joint motor output end and the load end can be obtained simultaneously, providing basic data for subsequent control.

[0066] Through the above technical solution, this invention can simultaneously acquire angle information from both the output end and the load end of the robot joint motor, providing the necessary data foundation for precise control of the robot joint motor. Furthermore, by comparing the first angle signal and the second angle signal, the elastic deformation in the transmission chain can be monitored in real time, effectively overcoming the interference of transmission chain elastic deformation on control accuracy. Therefore, this invention significantly improves the trajectory tracking stability of the robot joint under variable load conditions, achieves synchronous adaptation of the robot joint main motor torque command to load disturbances, and effectively suppresses end-effector trajectory deviation.

[0067] The present invention further proposes a method for performing hysteresis compensation on the second angle signal, including setting a dynamically adjustable delay time window for the compensation torque command of the robot joint auxiliary motor, wherein the reference value of the time window is 0.8 to 1.5 times the closed-loop control cycle of the robot joint main motor position; obtaining the delay coefficient by querying a preset mapping table according to the real-time load rate of the robot joint and adjusting the delay coefficient based on the temperature of the robot joint main motor; and applying the delay time window multiplied by the delay coefficient to the second angle signal.

[0068] The base value of the delay time window is set to 0.8 to 1.5 times the main motor control cycle, allowing the time window length to adaptively adjust according to the dynamic characteristics of the main motor control loop. The delay coefficient is obtained by querying a preset load rate mapping table; for example, the coefficient is set to 0.9 when the load rate is 0% to 30%, 1.0 when it is 30% to 70%, and 1.2 when it is 70% to 100%. A temperature adjustment mechanism further optimizes the delay coefficient. When the main motor temperature exceeds 65℃, an additional coefficient value of 0.05 to 0.1 is added based on the load rate. When the temperature exceeds 85℃, the coefficient is forcibly increased to 1.3 times the original value. The delay coefficient is updated at a frequency of 10 to 50 times per second to ensure dynamic response speed.

[0069] Specifically, the reference value of the delay time window is correlated with the main motor control cycle to keep the compensation delay synchronized with the system control frequency, avoiding phase mismatch. The load rate mapping table associates different load ranges with corresponding delay attenuation or enhancement coefficients. For example, the compensation intensity is reduced under low load to prevent high-frequency resonance, and the compensation intensity is increased under high load to offset elastic deformation. The temperature adjustment mechanism introduces thermal state parameters. When the motor temperature rises, causing a decrease in mechanical stiffness, the compensation action time is extended by increasing the delay coefficient to compensate for the transmission chain deformation lag caused by thermal expansion. The high-frequency update of the delay coefficient allows the compensation parameters to track load fluctuations and temperature changes in real time, forming a dynamic closed-loop adjustment. By establishing a multi-dimensional correlation between load rate, temperature, and delay coefficient, the compensation intensity can adapt to different operating condition combinations, effectively suppressing the accumulation of resonant energy, reducing servo energy consumption, and extending the reducer's lifespan.

[0070] As a preferred embodiment, the solution of the present invention is specifically implemented as follows:

[0071] A dynamically adjustable delay time window is set for the robot joint auxiliary motor compensation torque command. The baseline value of this time window is 1.2 times the closed-loop control cycle of the robot joint main motor position. The delay coefficient is obtained by querying a preset mapping table based on the real-time load rate of the robot joint. When the robot joint load rate is 40%, the delay coefficient is set to 1.0. Furthermore, the delay coefficient is adjusted based on the temperature of the robot joint main motor. For example, when the main motor temperature is 75℃, the delay coefficient is increased by an additional 0.1. Finally, the delay time window is multiplied by the adjusted delay coefficient to obtain 1.32 times the main motor position closed-loop control cycle as the final delay time, which is applied to the second angle signal.

[0072] Through the above technical solution, this invention achieves dynamic adjustment of the compensation torque command for the auxiliary motors of robot joints. Therefore, the hysteresis compensation operation can be adaptively adjusted according to real-time load and temperature changes, avoiding insufficient or excessive compensation problems caused by fixed delay parameters. Furthermore, by dynamically adjusting the delay coefficient, the generation of high-frequency and low-frequency resonances is effectively suppressed, improving the stability and accuracy of the robot joints. Specifically, under low-load and low-temperature conditions, the compensation intensity is reduced to avoid overcompensation; under high-load and high-temperature conditions, compensation is enhanced to counteract the accumulation of elastic deformation energy. This adaptive mechanism significantly improves the robot's performance under varying loads and temperature conditions.

[0073] The present invention further proposes the following logic for setting the delay coefficient: when the robot joint load rate is between 0% and 30%, the delay coefficient is set to 0.9; when the robot joint load rate is between 30% and 70%, the delay coefficient is set to 1.0; when the robot joint load rate is between 70% and 100%, the delay coefficient is set to 1.2; and the delay coefficient is updated at a frequency of 10 to 50 times per second.

[0074] The delay coefficient is divided into three load rate ranges: a coefficient lower than the baseline value for the low load range and a coefficient higher than the baseline value for the high load range. The delay coefficient update frequency is limited to 10-50Hz, ensuring real-time matching of compensation intensity with load status through a high-frequency refresh mechanism. The load rate ranges are divided using asymmetric boundaries, with 30% and 70% as key dividing points, forming three compensation intensity gradients: low, medium, and high.

[0075] Specifically, in the load range of 0%-30%, the delay coefficient is reduced to 0.9 to decrease the compensation amount and avoid high-frequency resonance caused by excessive compensation under low load. In the load range of 30%-70%, the baseline coefficient is maintained at 1.0 to ensure that the compensation strength matches the standard operating conditions. When the load exceeds 70%, the coefficient is increased to 1.2 to enhance compensation and offset the accumulation of elastic deformation energy caused by high load. The delay coefficient is updated at a frequency of 10-50 times per second, which ensures a rapid response to changes in load conditions while avoiding overloading the controller's computing resources due to excessively high frequencies. For example, under conditions of rapidly fluctuating load, using a 50Hz update frequency can ensure that the delay coefficient is adjusted within 20ms, allowing the compensation strength to accurately follow load changes.

[0076] As a preferred embodiment, the solution of the present invention is specifically implemented as follows:

[0077] The logic for setting the delay coefficient is as follows: when the robot joint load rate is between 0% and 30%, the delay coefficient is set to 0.9; when the robot joint load rate is between 30% and 70%, the delay coefficient is set to 1.0; when the robot joint load rate is between 70% and 100%, the delay coefficient is set to 1.2; and the delay coefficient is updated at a frequency of 10 to 50 times per second.

[0078] Specifically, a lookup table method can be used to dynamically set the delay coefficient. For example, a mapping table containing load rate ranges and corresponding delay coefficients can be pre-established. During operation, the corresponding delay coefficient is obtained by querying this mapping table based on the real-time detected robot joint load rate. To ensure the real-time performance of the delay coefficient, the system will periodically update the delay coefficient at a frequency of 10 to 50 times per second.

[0079] As one possible implementation, a timer can be used to trigger the update of the delay coefficient. The timer period can be set to a value between 20 milliseconds and 100 milliseconds, corresponding to an update frequency of 10 to 50 times per second. Each time the timer is triggered, the system executes the following steps:

[0080] Obtain the real-time load rate of the current robot joints;

[0081] The corresponding latency coefficient is determined by querying the mapping table based on the load rate.

[0082] The new delay factor is applied to the lag compensation operation.

[0083] Through the above technical solution, this invention can dynamically adjust the delay coefficient according to the real-time load of the robot joint, thereby optimizing the hysteresis compensation effect. Under low load conditions, a smaller delay coefficient can avoid overcompensation; under high load conditions, a larger delay coefficient can provide sufficient compensation strength. Frequent updates to the delay coefficient can ensure that the compensation operation can respond to load changes in a timely manner, improving the dynamic adaptability of the system. This adaptive compensation mechanism helps to improve the control accuracy and stability of the robot joint, while reducing the risk of resonance.

[0084] The present invention further proposes to adjust the delay coefficient based on the temperature of the robot joint main motor, including: real-time monitoring of the temperature of the robot joint main motor; when the temperature exceeds 65°C but does not exceed 85°C, if the current robot joint load rate is greater than or equal to 50%, the delay coefficient will be increased by an additional 0.1; if the current robot joint load rate is less than 50%, the delay coefficient will be increased by an additional 0.05; when the temperature of the robot joint main motor exceeds 85°C, the delay coefficient under all robot joint load rates will be forcibly set to 1.3 times the original value.

[0085] Temperature monitoring utilizes a PT100 temperature sensor mounted on the stator winding of the main motor, sampling temperature data 20 times per second. The delay coefficient adjustment logic is implemented through a preset temperature-load rate two-dimensional correction table, which is stored in the joint controller's non-volatile memory. When the temperature is between 65℃ and 85℃, the controller determines the compensation increment based on the real-time load rate: the compensation increment is set to 0.1 when the load rate is ≥50%, and 0.05 when the load rate is <50%. When the temperature exceeds 85℃, the controller triggers a temperature protection mode, forcibly multiplying the current delay coefficient by a correction factor of 1.3.

[0086] Specifically, when the main motor temperature reaches 70℃ and the load rate is 60%, the original delay coefficient is 1.0. An additional 0.1 is added to make the final coefficient 1.1. When the temperature rises to 90℃, regardless of the load rate range, the delay coefficient is forcibly increased to 1.3 times the original value. Temperature monitoring data and load rate data are processed synchronously via timestamps to ensure the timing consistency of compensation parameter adjustments. In temperature protection mode, the corrected delay coefficient is applied to the time window calculation module of the hysteresis compensation algorithm, increasing the compensation signal delay time by 30%, thereby enhancing the ability to offset thermally induced deformation. This mechanism actively suppresses the accumulation of elastic deformation and blocks the thermoplastic deformation chain reaction by increasing the compensation strength when the reducer stiffness decreases due to high temperature.

[0087] As a preferred embodiment, the solution of the present invention is specifically implemented as follows:

[0088] Monitor the temperature of the robot joint main motor in real time. When the temperature exceeds 65℃ but does not exceed 85℃, perform the following operations: if the current robot joint load rate is greater than or equal to 50%, increase the delay coefficient by an additional 0.1; if the current robot joint load rate is less than 50%, increase the delay coefficient by an additional 0.05. When the temperature of the robot joint main motor exceeds 85℃, forcibly set the delay coefficient of all robot joints under load rate to 1.3 times the original value.

[0089] For example, in one specific embodiment, the temperature of the robot joint's main motor is monitored to be 75°C. At this time, the system detects that the current robot joint load rate is 60%. Since the temperature exceeds 65°C but does not exceed 85°C, and the load rate is greater than or equal to 50%, the delay factor is increased by an additional 0.1. Assuming the original delay factor is 1.0, the adjusted delay factor is 1.1.

[0090] Furthermore, if the temperature of the robot joint's main motor continues to rise to 90°C, exceeding the 85°C threshold, the system will force the delay coefficient to be set to 1.3 times the original value, regardless of the current robot joint load rate. For example, if the original delay coefficient was 1.0, the adjusted delay coefficient would be 1.3.

[0091] Through the above technical solution, this invention achieves dynamic delay coefficient adjustment based on the temperature of the robot joint's main motor. Therefore, under high-temperature conditions, increasing the delay coefficient enhances the compensation strength, effectively offsetting the accumulation of elastic deformation energy caused by temperature rise. Specifically, when the temperature is between 65℃ and 85℃, different degrees of delay coefficient increase are adopted according to different load rates, achieving more precise compensation adjustment. When the temperature exceeds 85℃, a more aggressive delay coefficient increase strategy is adopted to cope with potentially more severe thermal effects. This temperature-related compensation enhancement mechanism effectively blocks the risk of plastic deformation caused by thermal effects, improving the stability and reliability of the robot joint.

[0092] This invention further proposes a delay coefficient as a proportional factor used to adjust the length of the delay time applied to the second angle signal of the robot joint during hysteresis compensation operation.

[0093] The delay coefficient, acting as a scaling factor, directly determines the final calculated value of the delay time window. Its value, multiplied by the time window, generates the actual compensation delay applied to the second angle signal. This coefficient establishes a correspondence with the load rate through a preset mapping table and is adjusted incrementally when there are temperature anomalies, forming a dynamic scaling mechanism. For example, when the load rate is 50% and the temperature is within acceptable limits, the delay coefficient is 1.0; if the temperature exceeds 65℃ and the load rate is greater than 50%, the coefficient increases to 1.1.

[0094] Specifically, in the hysteresis compensation operation, the baseline value of the delay time window is 0.8 to 1.5 times the main motor control cycle, and the delay coefficient is dynamically adjusted based on load rate range division and temperature monitoring. For example, when the load rate is in the 30%-70% range, the base coefficient is set to 1.0; if the main motor temperature exceeds 65℃, an increment of 0.05 or 0.1 is added according to the load rate. After multiplying the delay coefficient by the time window, the final delay time is generated and applied to the second angle signal. This proportional factor is quantitatively defined to ensure that the compensation delay time can accurately match load and temperature changes, avoiding over- or under-compensation caused by fixed parameters. For example, under high temperature and high load conditions, the delay coefficient is forcibly increased to 1.3 times the original value, significantly increasing the delay time to offset the effects of thermal deformation, block the accumulation of elastic deformation energy, thereby suppressing low-frequency resonance and reducing the risk of fatigue damage to the reducer.

[0095] As a preferred embodiment, the solution of the present invention is specifically implemented as follows:

[0096] The delay coefficient is a proportional factor used to adjust the duration of the delay applied to the second angle signal of the robot joint during hysteresis compensation. Specifically, the delay coefficient is a dimensionless value multiplied by a preset baseline delay time to obtain the actual delay time. For example, when the baseline delay time is set to 10 milliseconds, a delay coefficient of 0.9 results in an actual delay time of 9 milliseconds; a delay coefficient of 1.2 results in an actual delay time of 12 milliseconds. The delay coefficient can be dynamically adjusted based on the real-time load rate of the robot joint and the temperature of the main motor to adapt to compensation requirements under different operating conditions.

[0097] Through the above technical solution, this invention achieves flexible adjustment of hysteresis compensation operation during the coordinated drive of robot joint motors. Therefore, the delay time of the second angle signal can be precisely controlled according to actual working conditions, avoiding insufficient or excessive compensation caused by fixed delay parameters. Furthermore, this dynamic adjustment mechanism can effectively suppress system resonance, improve the control accuracy and stability of robot joints, while reducing servo energy consumption and extending the lifespan of key components.

[0098] The present invention further proposes setting a preset offset threshold, which includes dividing the robot joint load rate range and associating it with the threshold range. When the real-time load rate of the robot joint is less than or equal to 30%, the offset threshold range is set to 0.7° to 0.8°; when the real-time load rate of the robot joint is greater than 30% and less than or equal to 70%, the offset threshold range is set to 0.5° to 0.6°; when the real-time load rate of the robot joint is greater than 70%, the offset threshold range is set to 0.3° to 0.4°.

[0099] The offset threshold range is divided into three numerical intervals corresponding to three load rate ranges: 0.7° to 0.8° for low load rate, 0.5° to 0.6° for medium load rate, and 0.3° to 0.4° for high load rate. The boundaries between the load rate intervals are set at 30% and 70%, and joint load rate data is acquired in real time through a pre-set load rate detection module. The threshold range is negatively correlated with the load rate; that is, the higher the load rate, the smaller the threshold range.

[0100] Specifically, when the robot joint is in a low load rate range, the end effector is less affected by external disturbances, allowing for a larger angular deviation threshold to avoid false triggering of compensation polarity switching. At this time, the threshold range is set to 0.7° to 0.8°, and the corresponding threshold parameter is called through the threshold selection module built into the main motor controller. When the load rate enters the medium load rate range, the end effector is more affected by the elastic deformation of the transmission chain, and the threshold range is adjusted to 0.5° to 0.6° to balance trajectory accuracy and system stability. When the load rate exceeds 70%, the elastic deformation of the reducer can accumulate rapidly, and the threshold range is compressed to 0.3° to 0.4°. More sensitive deviation detection promptly triggers the auxiliary motor's compensation torque polarity switching, preventing the accumulation of deformation energy. The dynamic adjustment of the threshold parameters is achieved through a preset load rate-threshold mapping table, which is stored in the non-volatile memory of the joint controller.

[0101] As a preferred embodiment, the solution of the present invention is specifically implemented as follows:

[0102] The preset offset threshold is the actual torsional angle deviation between the robot joint load output end and the robot motor output shaft, with a preset offset threshold range of 0.3° to 0.8°. Specifically, during the coordinated drive of the robot joint motors, a first angle signal from the robot motor output shaft and a second angle signal from the robot joint load output end are acquired by a first encoder and a second encoder, respectively. Further, the difference between the second angle signal and the first angle signal is calculated in real time and compared with the preset offset threshold. Therefore, when the absolute value of the difference exceeds the preset offset threshold, the output polarity of the robot joint auxiliary motor compensation torque control command is controlled according to the magnitude relationship between the second angle signal and the first angle signal. For example, the offset threshold can be set to 0.5°. When the calculated absolute value of the difference is greater than 0.5°, if the second angle signal is greater than the first angle signal, the output polarity of the robot joint auxiliary motor compensation torque control command is controlled to be negative; if the second angle signal is less than the first angle signal, the output polarity of the robot joint auxiliary motor compensation torque control command is controlled to be positive.

[0103] Through the above technical solution, this invention can accurately detect and compensate for the angular deviation between the load output end of a robot joint and the output shaft of the motor. By setting a reasonable offset threshold range, overcompensation or undercompensation can be effectively avoided. Furthermore, by controlling the polarity of the compensation torque according to the direction of the angular deviation, precise adjustment of the robot joint position can be achieved. This improves the positioning accuracy and motion stability of the robot joint and reduces trajectory deviations caused by elastic deformation of the transmission chain.

[0104] The present invention further proposes to divide the robot joint load rate range and associate it with a threshold range. When the real-time load rate of the robot joint is less than or equal to 30%, the offset threshold range is set to 0.7° to 0.8°; when the real-time load rate of the robot joint is greater than 30% and less than or equal to 70%, the offset threshold range is set to 0.5° to 0.6°; when the real-time load rate of the robot joint is greater than 70%, the offset threshold range is set to 0.3° to 0.4°.

[0105] The offset threshold range is divided into three consecutive intervals, each corresponding to a different load condition. The boundaries between the load rate intervals are set at 30% and 70%, representing light load, medium load, and heavy load states. The threshold range decreases in a stepwise manner as the load rate increases, dynamically adjusting within the range of 0.3° to 0.8°. The mapping relationship between the load rate and the threshold range is achieved through a preset parameter table.

[0106] Specifically, by monitoring the robot's joint load rate in real time, the corresponding threshold range is dynamically selected. When the load rate is below 30%, a larger threshold of 0.7° to 0.8° is used, allowing for a large angular deviation between the main and auxiliary motors without triggering polarity switching, thus avoiding high-frequency oscillations caused by overcompensation under low load. When the load rate rises to the 30%-70% range, the threshold drops to 0.5° to 0.6°. At this point, the transmission chain stiffness decreases but has not reached its limit, and moderately early compensation can balance energy consumption and stability. When the load rate exceeds 70%, a minimum threshold of 0.3° to 0.4° is used, intervening in compensation early under high stress conditions in the reducer, effectively preventing the accumulation of elastic deformation energy. This dynamic threshold mechanism, by matching the load rate and the degree of transmission chain stiffness attenuation, synchronizes the timing of the compensation torque polarity switching with the deformation characteristics of the mechanical system in real time, controlling the balance between resonance suppression and energy consumption optimization within the optimal range.

[0107] In a preferred embodiment, the method for setting the preset offset threshold includes: establishing a robot joint load rate monitoring module to collect the current signal of the torque sensor at the reducer output end in real time and convert it into a load rate percentage value; creating a load rate interval division unit to divide the load rate into three level intervals: low, medium, and high; configuring a threshold mapping module to call the threshold database of 0.7° to 0.8° when the load rate is detected to be in the low interval, the threshold database of 0.5° to 0.6° when it is in the medium interval, and the threshold database of 0.3° to 0.4° when it is in the high interval; and setting a threshold dynamic switching logic to perform smooth interpolation of the threshold with a transition period of 20 milliseconds when the load rate crosses the boundary of an adjacent interval to avoid control oscillation caused by step changes.

[0108] Through the above technical solution, this invention effectively solves the problem of sensitivity mismatch under variable load conditions caused by traditional fixed offset thresholds. Using a larger threshold in the low load range prevents the joint auxiliary motor from frequently switching torque polarity due to minute deformations, avoiding high-frequency oscillations; using a smaller threshold in the high load range allows for timely detection of reducer elastic deformation, preventing deformation energy accumulation that could lead to low-frequency resonance. This dynamic threshold mechanism enables the system to automatically match the optimal deviation detection sensitivity under different load conditions, ensuring end-effector trajectory tracking accuracy while avoiding energy loss caused by overcompensation, significantly improving the robustness and energy efficiency of robot joint control.

[0109] The present invention further proposes a technical solution for dynamically associating the offset threshold range with the real-time load rate, including: dividing the robot joint load rate range and associating it with the threshold range; when the robot joint real-time load rate is less than or equal to 30%, the offset threshold range is set to 0.7° to 0.8°; when the robot joint real-time load rate is greater than 30% and less than or equal to 70%, the offset threshold range is set to 0.5° to 0.6°; when the robot joint real-time load rate is greater than 70%, the offset threshold range is set to 0.3° to 0.4°.

[0110] The load rate range is divided based on the nonlinear decay characteristics of the reducer stiffness. A 30% load rate corresponds to the starting point of stiffness decrease, and a 70% load rate corresponds to the critical point of accelerated stiffness decay. The threshold range setting adopts reverse correlation logic, lowering the threshold as the load rate increases to ensure earlier compensation triggering under high load conditions. The threshold bandwidth is set to 0.1° to provide adjustment margin for the control algorithm and avoid oscillations caused by sudden threshold changes.

[0111] Specifically, in the load rate monitoring module, the ratio of the joint output torque to the rated torque is calculated in real time. After eliminating high-frequency noise through a third-order Butterworth filter, the load rate is divided into three operating ranges. When the load rate is in the 0-30% range, a wide threshold band of 0.7-0.8° is used, allowing the drivetrain to produce moderate elastic deformation without triggering compensation, thus avoiding oversensitivity under low load. When the load rate enters the 70-100% range, a narrow threshold band of 0.3-0.4° is used, intervening in compensation in advance when the reducer stiffness decreases significantly, preventing the accumulation of deformation energy. The threshold switching process uses a hysteresis control algorithm. When the load rate fluctuates near the range boundary, the current threshold range is maintained until the load rate change exceeds the hysteresis width of 5%, preventing frequent switching. This scheme ensures that the compensation triggering timing is precisely matched with the actual stiffness characteristics of the drivetrain, maintaining optimal control response across the entire load rate range.

[0112] As a preferred embodiment, the solution of the present invention is implemented as follows: During the generation of the reference torque control command for the main motor of the robot joint, a preset load disturbance model is deployed in the real-time computing unit of the servo controller. This model receives the angle deviation signal θ from the first encoder. error This deviation value is the difference between the actual angle of the current motor output shaft and the theoretical target angle. The model also incorporates real-time joint load rate parameters provided by the joint load rate monitoring module. Mload This parameter is obtained by fusing current loop sampling with a Kalman filter. The stiffness attenuation factor ω is generated by converting temperature data collected by the main motor temperature sensor through a linear mapping table. When the main motor temperature is between 25℃ and 80℃, the value of ω ranges from 0.85 to 1.15. When the model performs decoupling operations, θ... error With M load The product operation is performed, and then the result is divided by the stiffness attenuation factor ω. Finally, the reference torque command τ of the main motor is output. main The torque command is transmitted to the main motor controller via the CAN bus at a refresh rate of 500Hz, enabling active decoupling compensation of the elastic deformation of the drive train.

[0113] Through the above technical solution, this invention effectively solves the trajectory deviation problem caused by the inability to decouple transmission chain deformation in traditional single encoder systems. By employing a three-dimensional decoupling mechanism that integrates angle deviation, load status, and temperature parameters in real time, the main motor torque command can actively counteract the reverse torque generated by the elastic deformation of the reducer. This solution maintains the stability of the main control loop while suppressing the trajectory deviation of the end effector within ±0.2°. Simultaneously, through the temperature compensation function of the stiffness attenuation factor, it avoids compensation inaccuracies caused by the decrease in material stiffness under high-temperature conditions.

[0114] This invention further proposes a method for dynamically decoupling the signal mismatch between the two encoders when generating the reference torque control command for the main motor of the robot joint. This robot joint load disturbance model satisfies the following formula: ;wherein τ main This is the reference torque command for the main motor of the robot joint; where θ error M represents the deviation between the first angle signal and the target angle, where the target angle is the theoretically expected angular position of the robot joint motor output shaft at the current moment; where M... load ω represents the real-time joint load rate of the robot joint; where ω is the stiffness attenuation factor of the robot joint's main motor temperature mapping.

[0115] The robot joint load disturbance model comprises a combined structure of proportional-integral-derivative (PID) control terms and load-temperature compensation terms. The PID control term consists of the instantaneous value of the angle deviation, a derivative term, and an integral term. The load-temperature compensation term is a nonlinear function of the joint load rate and the stiffness attenuation factor. The stiffness attenuation factor is obtained in real-time through experimentally calibrated temperature-stiffness curve mapping; for every 10°C increase in temperature, the stiffness attenuation factor decreases by 0.05-0.12. The calculation process of the load-temperature compensation term includes the following steps: first, the real-time joint load rate is input into a second-order polynomial function to calculate the basic compensation amount; then, the stiffness attenuation factor is applied as a gain coefficient to the basic compensation amount; finally, a limiter constrains the compensation amount within a preset range.

[0116] Specifically, during the closed-loop position control of the robot joint's main motor controller, motor temperature data is collected in real time, and the corresponding stiffness attenuation factor is calculated. When the motor temperature rises from 25℃ to 75℃, the stiffness attenuation factor linearly decreases from 1.0 to 0.82. At this point, the output of the load-temperature compensation term automatically increases by 22% to offset the impact of the decrease in transmission chain stiffness. In scenarios with sudden load rate changes, such as a jump from 30% to 80%, the second-order polynomial function causes an overshoot response in the basic compensation amount. The compensation amount increases to 1.3 times the steady-state value within 20ms, and then decays to the equilibrium value within 100ms. This dynamic compensation process works in conjunction with the proportional-integral-derivative (PID) control term, enabling the main motor torque command to generate over-torque output at the initial stage of load changes, effectively suppressing the accumulation of transmission chain deformation. Experimental data shows that this composite control model can reduce the end-effector trajectory deviation to ±0.12° while reducing servo energy consumption by 15%.

[0117] As a preferred embodiment, the present invention is implemented as follows: During the generation of the robot joint auxiliary motor compensation torque control command, the delay coefficient is defined as a proportional factor that adjusts the length of the delay time applied to the second angle signal in the hysteresis compensation operation. This proportional factor is initialized by acquiring joint load rate data in real time and querying a preset load rate-delay coefficient mapping table. Specifically, when the joint load rate is in the range of 0% to 30%, the initial value of the proportional factor is set to 0.9; when the load rate rises to the range of 30% to 70%, the proportional factor is adjusted to 1.0; if the load rate exceeds 70%, the proportional factor is increased to 1.2. Further, by monitoring the main motor temperature data in real time, when the temperature exceeds 65°C but does not reach 85°C, if the current load rate is higher than 50%, the proportional factor is increased by 0.1 based on the initial value; if the load rate is lower than 50%, it is increased by 0.05. When the main motor temperature exceeds 85°C, regardless of the load rate range, all proportional factors are forcibly adjusted to 1.3 times the initial value. The scaling factor is updated at a frequency of 10 to 50 times per second to ensure real-time dynamic adjustment.

[0118] Through the above technical solution, this invention achieves precise dynamic adjustment of the delay time in hysteresis compensation operation. By establishing a mapping relationship between the delay coefficient and load rate and temperature parameters, the problem of resonance caused by traditional fixed delay compensators under varying operating conditions is effectively solved. The compensation intensity is reduced under low load and low temperature conditions to avoid high-frequency oscillation, while the compensation is strengthened under high load and high temperature conditions to suppress low-frequency resonance. Simultaneously, a temperature-triggered forced compensation mechanism blocks the risk of thermal deformation. This solution significantly improves the stability of the closed-loop control of the transmission chain and extends the service life of key components such as the reducer.

[0119] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A robot joint motor cooperative driving method based on dual-encoder dynamic decoupling, characterized in that, The method comprises the following steps: Step 1: Real-time acquisition of a first angle signal of a robot motor output shaft through a first encoder installed on the robot motor output shaft; Step 2: Real-time acquisition of a second angle signal through a second encoder installed on a robot joint load output end; Step 3: Generation of a robot joint main motor reference torque control instruction according to the first angle signal and transmission to a robot joint main motor controller; Step 4: Hysteresis compensation operation on the second angle signal to generate a hysteresis compensation signal; Step 5: Generation of a robot joint auxiliary motor compensation torque control instruction based on the hysteresis compensation signal and transmission to a robot joint auxiliary motor controller; Step 6: Real-time calculation of the difference between the second angle signal and the first angle signal, and when the absolute value of the difference exceeds a preset offset threshold value: If the second angle signal is greater than the first angle signal, the output polarity of the robot joint auxiliary motor compensation torque control instruction is negative; If the second angle signal is less than the first angle signal, the output polarity of the robot joint auxiliary motor compensation torque control instruction is positive.

2. The robot joint motor cooperative driving method based on dual-encoder dynamic decoupling according to claim 1, characterized in that, The first angle signal is the real-time rotation angle of the robot motor output shaft, and the second angle signal is the actual position angle of the robot joint end effector.

3. The robot joint motor cooperative driving method based on dual-encoder dynamic decoupling according to claim 1, characterized in that, The hysteresis compensation operation in step 4 comprises: Step 4-1: Setting a dynamically adjustable delay time window for the robot joint auxiliary motor compensation torque instruction, and the reference value of the time window is 0.8 to 1.5 times the robot joint main motor position closed-loop control period; Step 4-2: Obtaining a delay coefficient from a preset mapping table according to the real-time load rate of the robot joint and adjusting the delay coefficient based on the temperature of the robot joint main motor; Step 4-3: Multiplying the delay time window and the delay coefficient and then acting on the second angle signal.

4. The robot joint motor cooperative driving method based on dual-encoder dynamic decoupling according to claim 3, characterized in that, The setting logic of the delay coefficient in step 4-2 is: When the robot joint load rate is 0% to 30%, the delay coefficient is set to 0.9; When the robot joint load rate is 30% to 70%, the delay coefficient is set to 1.0; When the robot joint load rate is 70% to 100%, the delay coefficient is set to 1.2; The delay coefficient is updated at a frequency of 10 to 50 times per second.

5. The robot joint motor cooperative driving method based on dual-encoder dynamic decoupling according to claim 3, characterized in that, The adjustment of the delay coefficient based on the temperature of the robot joint main motor in step 4-2 comprises: Step 4-2-1: Real-time monitoring of the temperature of the robot joint main motor, and when the temperature exceeds 65℃ but does not exceed 85℃: If the current robot joint load rate is greater than or equal to 50%, the delay coefficient is additionally increased by 0.1; If the current robot joint load rate is less than 50%, the delay coefficient is additionally increased by 0.05; When the temperature of the robot joint main motor exceeds 85℃, the delay coefficient under all robot joint load rates is forcibly set to 1.3 times the original value.

6. The robot joint motor cooperative driving method based on dual-encoder dynamic decoupling according to any one of claims 3-5, characterized in that, The delay coefficient is a proportional factor used to adjust the length of the delay time applied to the second angle signal of the robot joint in the hysteresis compensation operation.

7. The robot joint motor cooperative driving method based on dual-encoder dynamic decoupling according to claim 1, characterized in that, The preset offset threshold value in step 6 is the actual torsion angle deviation of the robot joint load output end relative to the robot motor output shaft, and the preset offset threshold value range is 0.3° to 0.8°.

8. The robot joint motor cooperative driving method based on dual-encoder dynamic decoupling according to claim 7, characterized in that, The setting of the preset offset threshold in step 6 comprises: dividing the robot joint load rate interval and associating the threshold range, when the real-time load rate of the robot joint is less than or equal to 30%, the offset threshold range is set to 0.7° to 0.8°; when the real-time load rate of the robot joint is > 30% and ≤ 70%, the offset threshold range is set to 0.5° to 0.6°; when the real-time load rate of the robot joint is greater than 70%, the offset threshold range is set to 0.3° to 0.4°.

9. The robot joint motor cooperative driving method based on dual-encoder dynamic decoupling according to claim 1, characterized in that, In the step 3, when the robot joint main motor reference torque control instruction is generated according to the first angle signal, the preset robot joint load disturbance model is used to dynamically decouple the double-encoder signal mismatch amount, and the robot joint load disturbance model satisfies the following formula: ; where τ main is the robot joint master motor reference torque command; where θ error is the deviation of the first angle signal from a target angle, the target angle being the angle position that the output shaft of the joint motor of the robot should reach at the current time instant; where M load is the real-time joint load ratio of the robot joint; Where ω is the stiffness attenuation factor of the robot joint main motor temperature mapping.

Images