Control method of joint module, joint module and robot

Abstract

The application belongs to the technical field of robot motion control, and particularly relates to a control method of a joint module, a joint module and a robot. When the joint module is in a zero-speed torque maintaining (stall) mode, a current basic torque is obtained, a sine disturbance current with a specific amplitude, frequency and phase is adaptively calculated and injected to a current loop given value. The three-phase winding local overheating problem caused by the absolute static state of the motor electric angle when the motor is stalled is broken. By injecting the sine disturbance current, the electric angle is forced to scan smoothly around the current working point, so that the three-phase winding of the motor alternately bears the peak current. Without increasing any additional heat dissipation hardware, the dynamic balanced distribution of the system heat is realized, the thermal balance duration of the robot in the large load static state is greatly prolonged, and the device damage caused by the local temperature rise exceeding the standard is effectively avoided.

Description

Technical Field

[0001] This invention belongs to the field of robot motion control technology, specifically relating to a control method for a joint module, a joint module, and a robot. Background Technology

[0002] In high-end servo control fields such as quadrupedal robot dogs, humanoid robots, and industrial robots, robots often need to be in a "static force exertion" state. For example, a robot may stand in place, maintain a specific posture, or grasp a load without making macroscopic displacement. From the perspective of motor control, the joint motor is in a "locked" condition where the rotor mechanical speed is zero, that is, the speed is zero, but it continuously outputs static torque.

[0003] In existing field-oriented control (FOC) algorithms for permanent magnet synchronous motors (PMSMs), when the motor rotates normally, the three-phase winding currents are alternating, uniform sine waves, resulting in relatively uniform heat generation. However, under stall conditions, the rotor's mechanical angle remains stationary, causing the electrical angle under FOC control to be fixed, and the instantaneous values ​​of the three-phase currents also remain unchanged. This results in a certain phase winding and its corresponding MOSFET continuously carrying extremely high peak currents. This extreme current unevenness leads to highly concentrated heat, causing the single-phase stator winding to heat up rapidly. This not only severely disrupts the system's thermal balance, limiting the robot's maximum load-bearing capacity and static time, but also easily leads to excessive local temperature rises and even burnout of the motor and driver. Summary of the Invention

[0004] In view of this, the present invention proposes a control method for a joint module, a joint module and a robot, aiming to improve the thermal balance effect and the maximum dwell time of the joint module.

[0005] In a first aspect, in a control method for a joint module provided by the present invention, the control method includes:

[0006] S100: Receives the basic torque current required from the host computer to put the joint module into zero-speed torque holding mode. I q_ref ;

[0007] S200, Generates sinusoidal disturbance current ;in, y (t) For the sinusoidal disturbance current generated in real time, A The magnitude of the disturbance current. f For the perturbation frequency, This is the phase compensation value;

[0008] S300, in the basic torque current I q_ref The sinusoidal disturbance current is superimposed on the above to obtain the reconstructed target torque current.I q * = I q_ref + y (t);

[0009] S400, the target torque current I q * The input current loop is closed-loop regulated, and the electrical angle of the motor driving the joint module is reciprocated smoothly. Within one scanning cycle, the peak current output by the three-phase inverter traverses the U, V, and W phase bridge arms.

[0010] Technical Effects: When the joint module is in zero-speed torque holding (locked rotor) mode, the current base torque current is acquired, and a sinusoidal disturbance current with specific amplitude, frequency, and phase is adaptively calculated and injected and superimposed onto the current loop setpoint. This overcomes the problem of localized overheating of the three-phase windings caused by the absolute stillness of the electrical angle when the motor is locked. By injecting a sinusoidal disturbance current, the drive electrical angle is forced to smoothly reciprocate near the current operating point, allowing the three-phase windings of the motor (and the driver MOSFETs) to alternately bear the peak current. Dynamic and balanced heat distribution of the system can be achieved without adding any additional heat dissipation hardware, significantly extending the thermal equilibrium duration (limit dwell time) of the robot under heavy load static conditions, and effectively avoiding device damage caused by excessive localized temperature rise.

[0011] In a preferred embodiment of the control method provided by the present invention, when the electrical angle of the motor driving the joint module performs a reciprocating smooth scan, the electrical angle of the motor of the joint module performs a reciprocating smooth scan within the range of -90° to 90° with the initial angle of the zero-speed torque holding mode as the center.

[0012] Technical Effects: In a permanent magnet synchronous motor, the three-phase windings are spatially separated by 120° electrical angles. Scanning within the range of -90° to 90° (a total of 180°) of electrical angle is a "theoretically optimal solution." This ensures that regardless of which sector the initial stall angle of the motor falls in, the synthesized current vector can completely traverse and alternately align with the coordinate axes of the other phases. This forces the phase initially carrying the peak current to be completely unloaded, truly achieving absolute heat distribution among the U, V, and W phases.

[0013] In a preferred embodiment of the control method provided by the present invention, the time during which the power switching transistors of the U, V, and W phase bridge arms of the three-phase inverter are individually turned on to withstand the peak current is less than half of the scanning cycle.

[0014] Technical benefits: By limiting the peak current of a single-phase bridge arm (MOSFET) to less than half a cycle, the path of continuous heat accumulation is directly cut off from the physical heat generation level. This gives the power switching transistor sufficient heat dissipation recovery time, effectively preventing the junction temperature of a single MOSFET from exceeding the standard or causing the device to explode due to continuous "full load heating", fundamentally improving the extreme survivability and lifespan of the driver hardware.

[0015] In a preferred embodiment of the control method provided by the present invention, the disturbance current amplitude A The number of pole pairs of the motor of the joint module p Negative correlation ， And the amplitude of the disturbance current A The reduction ratio of the reducer of the joint module n It shows a negative correlation.

[0016] Technical effect: The ultimate goal of the disturbance is to produce a sufficient change in the "electric angle" (e.g., 180°). According to the principles of motors (electric angle = number of pole pairs), this is achieved by... p × Mechanical angle), number of pole pairs p The more motors there are, the more significant the changes in electrical angle that can be generated with only minimal rotor mechanical movement; similarly, the reduction ratio... p The larger the value, the smaller the mechanical displacement transmitted to the joint output end by the rotor's minute oscillations. Simultaneously, the current amplitude needs to be clearly defined. A and p , n The negative correlation means that when facing multi-pole logarithmic motors or large reduction ratio joints, the system will automatically "converge" and reduce the amplitude of the disturbance current. This ensures that, under the premise of meeting the electrical angle scanning range (thermal equilibrium), the output shaft will not produce excessively amplified harmful mechanical oscillations, thereby protecting the gears of harmonic / RV reducers from the impact of large-amplitude high-frequency alternating stress at the source, and ensuring the stability and smoothness of the joint module's overall operation.

[0017] In a preferred embodiment of the control method provided by the present invention, based on the fundamental torque current... I q_ref The adaptive proportional calculation of the disturbance current amplitude A : ,in, l ( p , n () is determined by the number of pole pairs of the motor. p and the reduction ratio of the reducer n The determined baseline scaling factor, l ( p , n This is a preset value within the empirical range of 0.05 to 0.15.

[0018] Technical effect: Reduces the amplitude of the disturbance A It is no longer a fixed value, but can dynamically change with the current load (base torque current) to ensure that the optimal heat dissipation target can be achieved under any load, allowing the electrical angle to scan about 180°.

[0019] In a preferred embodiment of the control method provided by the present invention, the amplitude of the disturbance current is calculated based on the equivalent torsional stiffness of the joint module. A : , K load For the stiffness of the reducer output end, K t is the torque constant of the motor.

[0020] Technical effect: Introducing a reduction ratio n and equivalent torsional stiffness ( The calculations fully consider the flexible deformation of the joint's mechanical structure under different forces. This effectively avoids excessive mechanical oscillation at the end or system resonance caused by excessively large injected disturbance amplitude being amplified by the reducer, thus protecting the mechanical transmission components.

[0021] In a preferred embodiment of the control method provided by the present invention, the disturbance current amplitude A Based on the multidimensional interpolation lookup table method of offline calibration, it is determined that during the factory calibration stage of the joint module, the pole-log number will be... and reduction ratio As a fixed value, with the base torque current I q_ref As an input index, a lookup table is built; by applying different base torque currents... I q_ref To record the corresponding disturbance current amplitude. A This is done to fill the grid point data in the lookup table; during joint module operation, the data is based on the real-time acquired base torque current. I q_ref Look up the table to get the corresponding disturbance current amplitude. A Among them, if the basic torque current I q_ref If the disturbance current amplitude is located between two grid points, a linear interpolation algorithm is used to smoothly calculate the corresponding disturbance current amplitude. A .

[0022] Technical benefits: The underlying MCU of the joint module's driver has limited computing power. By offline calibration and the establishment of a look-up table, complex floating-point or nonlinear calculations are transformed into table lookup operations, significantly reducing the computational delay of the current loop control and ensuring high-frequency response capability. For non-grid point data, a linear interpolation algorithm is used to ensure that the amplitude of the output disturbance current remains continuous and smooth even when the load changes continuously (i.e., the torque current changes dynamically), without any abrupt changes, thus avoiding current noise or jerking in the motor.

[0023] In a preferred embodiment of the control method provided by this invention, if the current working condition of the robot allows the mechanical swing amplitude α of the joint module to be ≥0.3°, then 1Hz ≤ f ≤ 50Hz; otherwise, if the robot's current operating conditions do not allow for mechanical oscillation of the joint modules, 50Hz < f ≤ 300Hz.

[0024] Technical Effects: By employing a dual-mode frequency decoupling strategy, a low frequency of 1Hz~50Hz is used when mechanical micro-motion is permissible, and a high frequency of 50Hz~300Hz is used when micro-motion is not permissible. This creatively decouples "thermal management" from "mechanical constraints." Using the low frequency in non-precision operations (such as a robot dog standing on a rough surface) achieves an extreme improvement in heat dissipation (extending thermal equilibrium time by approximately 50%). In precision assembly or scenarios requiring quiet operation, the high frequency utilizes mechanical inertia filtering (low-pass filtering effect) to attenuate vibrations to the micrometer level, rendering it imperceptible to the naked eye and inaudible to the human ear, while still achieving approximately 20% improvement in heat dissipation.

[0025] In a preferred embodiment of the control method provided by the present invention, the phase compensation value The initial value is set to 0, and the disturbance current amplitude is... A and disturbance frequency f Once determined, based on the basic torque current I q_ref The current loop control cycle is adjusted accordingly.

[0026] Technical effect: The driver injects sinusoidal disturbance current y At time (t), the hardware and software system and control link will have inherent transmission delays. This is addressed by dynamically adjusting the phase compensation value in real time based on the current loop period. This ensures that the injected sinusoidal disturbance waveform is precisely aligned and integrated with the basic FOC current waveform, preventing beat frequency or high-order harmonic interference caused by phase difference, thus guaranteeing the stability and smoothness of closed-loop control.

[0027] In a preferred embodiment of the control method provided by the present invention, when generating a sinusoidal disturbance current... y Before (t), it also includes determining whether the joint module is in zero-speed torque holding mode based on the following conditions: receiving an instruction from the host computer to enter zero-speed torque holding mode; or, detecting that the motor speed reaches the duration threshold within the preset zero-speed threshold range.

[0028] Technical effects: (1) By accepting instructions from the host computer, the robot can "actively predict" before performing heavy-load grasping or static actions and enter the heat equalization mode in advance to prevent heat accumulation at the start; (2) By relying on the bottom-level autonomous monitoring, a "safety bottom line" is formed. Even if the host computer communication is abnormal or no instructions are issued, the joint module can automatically trigger the thermal equalization protection mechanism when it detects an unexpected stall (such as jamming or obstruction), which greatly improves the fault tolerance and survivability of the system.

[0029] In a second aspect, in a joint module provided by the present invention, the joint module includes a motor, a reducer, an encoder, and a driver. The reducer is connected to the output shaft of the motor, the encoder is used to detect the rotational speed of the motor, and the driver is electrically connected to the motor and the encoder respectively. The driver is configured to perform the control method of the joint module described in any one of the first aspects.

[0030] Technical benefits: This invention provides a highly integrated intelligent joint module with "autonomous thermal management capabilities." This joint module eliminates the need for external water-cooling housings, cooling fans, or additional temperature sensors. Through upgrades to the driver software algorithm, it overcomes physical heat dissipation bottlenecks at extremely low hardware costs, achieving a higher rated holding torque density.

[0031] Thirdly, in a robot provided by the present invention, the robot includes the joint module described in the second aspect above.

[0032] Technical effects: It endows robots (especially legged robot dogs and humanoid robots) with superior extreme load-bearing capacity and extended static working time. It significantly optimizes the overall NVH (noise, vibration, and harshness) performance of robots under extreme working conditions, improving the robot's environmental adaptability, operational accuracy, and overall service life. Attached Figure Description

[0033] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which will make the above and other features and advantages of the present invention more apparent to those skilled in the art. In the drawings:

[0034] Figure 1 This is a schematic diagram of the core flow of the control method for the joint module in this embodiment.

[0035] Figure 2 This is a schematic diagram illustrating the specific flow of the control method for the joint module in this embodiment. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0037] I. Hardware System Architecture and Overall Control Principles

[0038] In this embodiment, a robot is provided, including multiple joint modules. Each joint module specifically includes: a motor, a reducer (such as a harmonic reducer or RV reducer) connected to the motor's output shaft, an encoder for detecting the motor's speed and position, and a servo driver electrically connected to both the motor and the encoder. The joint module control method in this embodiment also relies on a robot host computer responsible for sending macroscopic motion commands.

[0039] When robots (such as legged robots and collaborative robots) are performing tasks such as standing still, maintaining posture, or bearing weight, the joint modules do not undergo macroscopic displacement, thus entering a "zero-speed torque holding mode" (i.e., stall condition). In traditional FOC (field-oriented control), stall would cause electrical angle... i Fixed, according to I a = I peak *cos( i According to the vector mapping formula, the three-phase current will be "frozen", causing a certain phase winding and the corresponding inverter bridge arm (MOSFET) to be subjected to a very large constant peak current and overheat rapidly.

[0040] The driver in this embodiment runs an optimized FOC algorithm. Its core idea is to actively inject a sinusoidal disturbance current with specific parameters into the current loop without changing the robot's macroscopic static posture, forcing the electrical angle to generate reciprocating micro-motion, so that the U, V, and W three-phase windings alternately bear large current, thereby achieving overall heat dissipation.

[0041] The core control methods of this joint module include:

[0042] S100: Receives the basic torque current required from the host computer to put the joint module into zero-speed torque holding mode. I q_ref ;

[0043] S200, Generates sinusoidal disturbance current ;in, y (t) For the sinusoidal disturbance current generated in real time, A The magnitude of the disturbance current. f For the perturbation frequency, This is the phase compensation value;

[0044] S300, in the basic torque current I q_ref The sinusoidal disturbance current is superimposed on the above to obtain the reconstructed target torque current. I q * = I q_ref + y (t);

[0045] S400, the target torque current I q * The input current loop is closed-loop regulated, and the electrical angle of the motor driving the joint module is reciprocated smoothly. Within one scanning cycle, the peak current output by the three-phase inverter traverses the U, V, and W phase bridge arms.

[0046] When the joint module is in zero-speed torque holding (locked rotor) mode, the current base torque current is acquired, and a sinusoidal disturbance current with specific amplitude, frequency, and phase is adaptively calculated and injected and superimposed onto the current loop setpoint. This overcomes the problem of localized overheating of the three-phase windings caused by the absolute stillness of the electrical angle when the motor is locked. By injecting a sinusoidal disturbance current, the drive electrical angle is forced to smoothly reciprocate near the current operating point, allowing the three-phase windings of the motor (and the driver MOSFETs) to alternately bear the peak current. Without adding any additional heat dissipation hardware, dynamic and balanced heat distribution of the system can be achieved, significantly extending the thermal equilibrium duration (limit dwell time) of the robot under heavy load static conditions, effectively avoiding device damage caused by excessive localized temperature rise.

[0047] II. Specific Implementation Steps of the Control Method

[0048] The core of the joint module control method in this embodiment lies in actively injecting a sinusoidal disturbance into the FOC current loop to force the three phases to alternately generate heat. The complete process includes the following steps:

[0049] Step S1: Triggering and predicting entry into zero-speed torque holding mode

[0050] To balance active feedforward and passive protection, the system provides a dual triggering mechanism:

[0051] 1. Host computer command trigger: The robot controller (host computer) actively sends an enable command to the driver to "enter zero speed torque holding mode (heating mode)" based on the current kinematic state (such as about to enter heavy load grasping or static standing).

[0052] 2. Low-level autonomous triggering: If no command is received from the host computer, the driver autonomously monitors the motor speed. When the detected speed is within the preset zero-speed threshold range (e.g., ...), the driver will trigger the motor. When the duration threshold (e.g., 1 second) is reached within a certain time, it is determined to be an unexpected jam or stall, and this mode is automatically triggered.

[0053] This dual-trigger mechanism creates an extremely high system fault tolerance. The host computer trigger enables "proactive feedforward," intervening before heat buildup occurs; the underlying trigger acts as a "safety bottom line," automatically activating thermal protection in case of unexpected blockages, even in the event of communication anomalies, significantly improving the safety and reliability of industrial applications.

[0054] Step S2: Adaptive generation of sinusoidal disturbance current parameters

[0055] The driver obtains the current current loop to maintain the base torque current required to sustain the external load. I q_ref And dynamically generate sinusoidal disturbance current. The three core parameters:

[0056] 1. Disturbance current amplitude A Adaptive computation:

[0057] Disturbance current amplitude A It is not a fixed value, but rather a setting to ensure that the motor's electrical angle can perform a precise, smooth reciprocating scan within the range of -90° to 90° from the initial stall angle. Disturbance current amplitude. A The setting rule is: the number of pole pairs of the motor p It is negatively correlated with the reduction ratio of the reducer. n The correlation is negative. Two implementation methods are provided:

[0058] Method 1 (Proportional Scaling Method): .in l ( p , n () is the base scaling factor, for different p and n , l The preset empirical value range is between 0.05 and 0.15. Basic torque current. I q_ref The larger the value, the greater the injected disturbance current.

[0059] Method 2 (Stiffness Model Method): .in K load This is the equivalent torsional stiffness at the output end of the reducer. K t Let be the torque constant of the motor. This formula fully considers the flexibility of the mechanical transmission chain and accurately calculates the current that just breaks static friction without causing excessive mechanical deformation.

[0060] In addition, table lookup optimization is provided. To save MCU computing power, during the factory calibration phase, [the following is omitted as the text is incomplete and cannot be translated]. p and n Treated as a fixed value, I q_ref As an input index, the corresponding amplitude is calculated and recorded offline. A A lookup table is created. At runtime, the driver directly uses real-time... I q_ref table lookup A ,like I q_ref If the points are located between grid points, a linear interpolation algorithm is used for smooth calculation.

[0061] In this embodiment, the disturbance amplitude Instead of using fixed, rigid values, this step ensures that the electrical angle produces a smooth scan of approximately 180° around the initial angle. This achieves a perfect match between "electromagnetic disturbance" and "mechanical response." Blindly applying a disturbance current could easily cause resonance in the robotic arm. This step ensures that a "just right" disturbance current is output under any load, guaranteeing the electrical angle scan range required for heat dissipation while avoiding gear fretting wear and system resonance caused by overdrive.

[0062] 2. Disturbance frequency f Dual-mode decoupling configuration:

[0063] To meet the constraints of different mechanical operating conditions, the host computer selects the frequency based on the operating condition flag:

[0064] Low-frequency oscillation mode: When the robot's operating conditions allow for slight mechanical oscillation of the joint modules (oscillation amplitude α satisfies α ≥ 0.3°), such as a quadruped robot dog standing on rough ground, a low frequency of 1Hz ≤ is set. f ≤ 50Hz. At this frequency, the heat dissipation effect is excellent, and the duration of thermal equilibrium can be increased by approximately 50%.

[0065] High-frequency silent mode: When the working conditions require absolute stillness or silence (no mechanical oscillation is allowed), such as when a collaborative robot is performing precision assembly, set the high frequency to 50Hz. f≤ 300Hz. At this point, the mechanical inertia of the reducer is used to generate a "low-pass filter" effect, which filters out micro-motions that are invisible to the naked eye. While avoiding the system's resonance point, it can still achieve a heat dissipation improvement of about 20%.

[0066] 3. Perturbation Phase Follow-up adjustment:

[0067] The initial value is set to 0. During operation, due to the periodic delay in digital control, the driver adjusts the current loop control period based on the base torque current. Perform advance follow-up adjustments to ensure that the disturbance waveform can be smoothly integrated with the underlying control loop and eliminate beat frequency.

[0068] In this step, the control is dynamically adjusted in real time according to the current loop control cycle. This offsets the inherent discrete delay of the digital system, ensures precise alignment between the disturbance waveform and the basic closed-loop current, eliminates the beat frequency phenomenon, and guarantees extremely high stability of the current loop.

[0069] Step S3: Injection of sinusoidal disturbance current and thermal equalization scan

[0070] The generated sinusoidal disturbance current y (t) Real-time, direct superposition to the FOC algorithm q The shaft current reference value, i.e., the final torque current setpoint, is... I q * = I q_ref + y (t).

[0071] After superposition, the FOC closed-loop control forces the inverter's output voltage vector to rotate in space. This causes the motor's electrical angle to perform a smooth reciprocating scan within the range of -90° to 90°, centered on the initial angle. During continuous scanning, the peak current of the three phases smoothly alternates between the U, V, and W phases as the angle changes. Because the scanning range reaches 180°, the time that any phase arm power switch (MOSFET) of the three-phase inverter can conduct alone and withstand peak current is forcibly controlled within half of the scanning cycle, completely eliminating local heat accumulation.

[0072] As the electrical angle continuously crosses, the U, V, and W phase windings take turns bearing the peak current. The time that any single phase bridge arm power transistor bears the peak current alone is forcibly compressed to less than half a cycle. This fundamentally breaks the static heat distribution during stall, evenly distributing the concentrated Joule heat from a single point to the entire stator and the three sets of MOSFETs, greatly improving the system's rated dwell torque without the need for any external heat dissipation hardware.

[0073] Step S4: Exit zero-speed torque holding mode

[0074] When the actuator receives an "Exit Heating Mode" command from the host computer, or detects that the rotation speed command from the main controller is not zero (the robot is preparing to perform a large-scale movement), the actuator internally uses a first-order low-pass filter to reduce the disturbance amplitude. A The torque decays smoothly to zero in a very short time (e.g., 10ms), thus seamlessly switching back to the standard FOC control mode. This avoids the torque jump that occurs during the control state switch and ensures an absolutely smooth transition of the robot from a static force exertion state to a macroscopic motion state.

[0075] III. Specific Application Scenarios and Examples

[0076] Scenario 1 (corresponding to low-frequency mode):

[0077] A hexapod robot is in a standby state on an uneven surface, allowing its end effector to slightly adapt to the terrain. At this time, the actuators are triggered to enter a warm-up mode, and the perturbation frequency is set. f = 5Hz. Based on the current high load of 20Nm (corresponding to a large base torque current). I q_ref The system adaptively calculates the amplitude of the disturbance current when the disturbance amplitude is large. A The electrical angle performs a sinusoidal reciprocating scan within a 180° range. Tests show that, due to the complete distribution of heat generation across the three phases, the thermal equilibrium time of the motor under heavy load at 25°C is significantly extended from the original 60 minutes to 90 minutes, improving the ultimate dwell time capability by 50%.

[0078] Scenario 2 (corresponding to high-frequency mode):

[0079] An industrial robotic arm grips a heavy object and hovers it at a precision inspection station, requiring its end effector to be absolutely still with no visible vibration. The system triggers a cooling mode and invokes a high-frequency strategy, setting the disturbance frequency. f = 150Hz, the mechanical resonance point of this arm type was avoided during debugging. At this time, although the stator magnetic field is scanning at high frequency inside, the high frequency vibration is filtered out by the inertia of the mechanical structure, and the amplitude of the end is controlled at the micrometer level, which meets the requirements of precision working conditions; and 150Hz does not produce harsh electromagnetic noise. In this mode, the thermal equilibrium time is still improved by 20%.

[0080] It should be noted that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0081] It should be understood that although this specification is described according to various embodiments, not every embodiment or implementation method contains only one independent technical solution. This way of describing the specification is only for clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0082] The above descriptions are merely illustrative embodiments of this application and are not intended to limit the scope of the embodiments of this application. Any equivalent changes, modifications, and combinations made by those skilled in the art without departing from the concept and principles of the embodiments of this application should fall within the protection scope of the embodiments of this application.

Claims

1. A control method for a joint module, characterized in that, The control method includes: Receives the base torque current required from the host computer to put the joint module into zero-speed torque holding mode. I q_ref ; Generate sinusoidal disturbance current ;in, y (t) represents the sinusoidal disturbance current generated in real time. A The magnitude of the disturbance current. f For the perturbation frequency, This is the phase compensation value; The basic torque current I q_ref The sinusoidal disturbance current is superimposed on the above to obtain the reconstructed target torque current. I q * = I q_ref + y (t); The target torque current I q * The input current loop is closed-loop regulated to drive the motor of the joint module to perform a smooth reciprocating scan of the electrical angle. Within one scan cycle, the peak current output by the three-phase inverter traverses the U, V, and W phase bridge arms.

2. The control method according to claim 1, characterized in that, When the electrical angle of the motor driving the joint module is smoothly reciprocated, the electrical angle of the motor of the joint module is smoothly reciprocated within a range of -90° to 90°, centered on the initial angle in the zero-speed torque holding mode; or, The time that the power switching transistors of the U, V, and W phase arms of the three-phase inverter are individually turned on to withstand peak current is less than half of the scanning cycle.

3. The control method according to claim 1, characterized in that, The amplitude of the disturbance current A The number of pole pairs of the motor of the joint module p Negative correlation ， And the amplitude of the disturbance current A The reduction ratio of the reducer of the joint module n It shows a negative correlation.

4. The control method according to claim 1, characterized in that, Based on the aforementioned basic torque current I q_ref The adaptive proportional calculation of the disturbance current amplitude A : ,in, λ ( p , n () is determined by the number of pole pairs of the motor. p and the reduction ratio of the reducer n The determined baseline scaling factor, λ ( p , n This is a preset value within the empirical range of 0.05 to 0.15; or, The amplitude of the disturbance current is calculated based on the equivalent torsional stiffness of the joint module. A : , K load For the stiffness of the reducer output end, K t is the torque constant of the motor.

5. The control method according to claim 1, characterized in that, The amplitude of the disturbance current A Determined using a multidimensional interpolation lookup table method based on offline calibration: During the factory calibration phase of the joint module, the pole number will be... and reduction ratio As a fixed value, with the base torque current I q_ref Use the input index to build a lookup table; By applying different base torque currents I q_ref To record the corresponding disturbance current amplitude. A This is to complete the filling of the grid point data in the lookup table; During joint module operation, based on the real-time acquired baseline torque current... I q_ref Look up the table to get the corresponding disturbance current amplitude. A Among them, if the basic torque current I q_ref If the disturbance current amplitude is located between two grid points, a linear interpolation algorithm is used to smoothly calculate the corresponding disturbance current amplitude. A .

6. The control method according to claim 1, characterized in that, If the robot's current operating conditions allow for a mechanical swing amplitude α ≥ 0.3° for the joint modules, then 1Hz ≤ f ≤50Hz; otherwise, If the robot's current operating conditions do not allow for mechanical oscillation of the joint modules, 50Hz < f ≤ 300Hz.

7. The control method according to claim 1, characterized in that, Phase compensation value The initial value is set to 0, and the disturbance current amplitude is... A and disturbance frequency f Once determined, based on the basic torque current I q_ref The current loop control cycle is adjusted accordingly.

8. The control method according to claim 1, characterized in that, In generating sinusoidal disturbance current y Before (t), it also includes determining whether the joint module is in zero-speed torque holding mode based on the following conditions: Received a command from the host computer to enter zero-speed torque holding mode; or, The motor speed was monitored to reach the duration threshold within the preset zero speed threshold range.

9. A joint module, characterized in that, include: Electric motor; A speed reducer, which is connected to the output shaft of the motor; An encoder, used to detect the rotational speed of the motor; A driver, which is electrically connected to the motor and the encoder respectively, is configured to perform the control method of the joint module as described in any one of claims 1 to 8.

10. A robot, characterized in that, Includes the joint module as described in claim 9.

Images