Motor control method for tail-end single-position sensor robot joint module
Through the fusion method of virtual axis model and position-free observer, the error and slow response problems of a single position sensor in the robot joint module are solved, and high-precision and high-dynamic motor control is achieved, reducing the module size and cost.
Patent Information
- Application Number
- CN202510539234.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-27
- Publication Date
- 2025-08-08
AI Technical Summary
When existing robot joint modules use only a single position sensor at the end of the reducer, there are problems such as large error in capturing the end position information and slow response speed, which makes it difficult to achieve high-precision control.
The motor control method of the end single position sensor is adopted. By establishing a virtual axis model and a positionless observer, combining high-frequency injection and back-potential observation, it combines inaccurate conversion angles and non-fast positionless observation angles to provide high dynamic and high-precision motor rotor position information.
It realizes that the robot joint module can achieve the same response speed and control accuracy as the dual-position sensor module when a single position sensor is installed at the end of the reducer, reducing the module size and cost.
Smart Images
Figure CN120454553A_ABST
Abstract
Description
Technical Field
[0001] The present invention is a control method using only a terminal single position sensor, and relates to the technical fields of robot joint modules and motor drives. Background Art
[0002] Currently, high-precision positioning control of robot joints relies on position sensors at the output of the joint's reducer. However, control of the joint motor itself relies on position sensors on the motor rotor. Consequently, most joints are equipped with dual position sensors, one located at the motor rotor and one at the reducer output. This significantly limits joint design flexibility and high integration. In some scenarios where precision is less critical, a single sensor on the motor rotor is used to control the motor and infer the position of the reducer end. However, this cannot achieve absolute positioning of the reducer output position for different rotor turns and the same angle. Therefore, it is desirable to install a single position sensor at the reducer end to achieve high-precision, high-response control of the entire joint. When using a single position sensor at the end, the motor angle is inaccurate due to factors such as backlash and amplification. Observation using a position-free control algorithm results in a slow, inaccurate change in the motor angle. Currently, effective control methods are lacking. Summary of the Invention
[0003] Technical Problem: This invention provides a motor control method for a robot joint module with a single-position sensor at the end, overcoming the errors and slow tracking response caused by capturing end position information and converting motor angles using a single-position sensor. This control method establishes a virtual axis to fuse the inaccurate converted angle with the slow, non-positioned observation angle, providing highly dynamic and high-precision motor rotor position information for motor drive control. This method overcomes the drawback of current robot joint modules requiring the installation of position sensors on the motor rotor, reduces the size and cost of the joint module, and ensures that its performance is consistent with the response speed and control accuracy of existing dual-position sensor modules.
[0004] Technical solution: In the motor control method of a robot joint module with a single position sensor at the end of the present invention, the joint module includes a motor drive controller, a motor, a reducer, a reducer output terminal, a position sensor, and a control algorithm module running in the main control chip of the motor drive controller circuit board; the output signal of the motor drive controller is output to the motor, and the motor current i is abc ) is output to the control algorithm module; the output end of the motor is connected to the reducer; a position sensor is set at the output end of the reducer to output the end position information θ0 to the control algorithm module; the control algorithm module outputs the control pulse width PWM signal to the motor drive controller to complete a closed-loop control system.
[0005] The control algorithm module collects system feedback information, including at least the motor current i abc The terminal position information θ0 collected by the position sensor; at the same time, the control algorithm module provides control information to the motor drive controller, including at least the control pulse width PWM signal required by the three-phase driver; the control algorithm module includes a position sensorless observation algorithm, specifically a high-frequency injection positionless observer, a back-EMF positionless observer, a reducer ratio conversion module, a virtual axis model, and a method for generating the angle information required for motor control.
[0006] Depending on the motor speed, the high-frequency injection position-free observer operates at low speed and the back-EMF position-free observer operates at medium and high speed. At the same time, only one of the observers is running. The specific operation is determined by the pre-defined speed threshold, and one of the two is selected by the high-speed switching to obtain the estimated position of the position-free observer. The estimated position, minus the angle information After the difference is calculated, the first virtual electromagnetic torque value is obtained through the control module K1(s):
[0007] The terminal position information θ0 is combined with the mechanical ratio of the reducer and the motor angle conversion value is obtained by the reducer ratio conversion module. The converted value, minus the angle information After the difference is calculated, the second virtual electromagnetic torque value is obtained through the control module K2(s):
[0008] The input of the virtual axis model is composed of three parts: the electromagnetic torque estimate calculated by the motor torque calculation module. The first virtual electromagnetic torque value The second virtual electromagnetic torque value The transfer function of the virtual axis model is Among them J m is the virtual moment of inertia of the virtual axis, s is the Laplace frequency variable, and its value is as close as possible to the real physical moment of inertia of the robot joint module; the output of the virtual axis model is angle information
[0009] The control module K1(s) adopts one of the following control methods: P control, i.e. proportional control; PI control, i.e. proportional integral control; PID control, i.e. proportional integral differential control.
[0010] The control module K2(s) adopts one of the following control methods: P control, i.e. proportional control; PI control, i.e. proportional integral control; PID control, i.e. proportional integral differential control.
[0011] The control algorithm module also uses a general motor joint module control module, including a motion control module, a filtering and separation module, a coordinate transformation module, a current control module, a coordinate inverse transformation module, an SVPWM module, and a high-frequency injection generator module.
[0012] To ensure complete system operation, the system also uses a universal motor joint module control module, including a motion control module, a filter separation module, a coordinate transformation module, a current control module, an inverse coordinate transformation module, an SVPWM module, and a high-frequency injection generator module. These control modules are well known in the art.
[0013] Beneficial effect: Compared with the prior art, in the motor control method of the end single-position sensor robot joint module of the present invention, the robot joint module only needs to install a single position sensor at the end of the reducer. Through the control method, the errors and slow tracking response caused by the single-position sensor in capturing the end position information and converting the motor angle are overcome, and the fusion of inaccurate conversion angles and slow positionless observation angles is realized, providing high-dynamic and high-precision motor rotor position information for motor drive control, reducing the volume and cost of the joint module, and the design complexity of the mechanical combination. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] Figure 1 It is the system composition and control block diagram of the present invention.
[0015] The figure includes: motor drive controller 1, motor 2, reducer 3, reducer output terminal 3.1, position sensor 4, control algorithm module 5, motor current i abc 6. End position information θ07. Control pulse width PWM signal 8. High-frequency injection without position observer 5.1. Back EMF without position observer 5.2. Reducer ratio conversion module 5.3. Virtual axis model 5.4. High and low speed switching selection 5.5. Estimated position Motor angle conversion value Control module K1(s) 5.8, first virtual electromagnetic torque value Control module K2(s) 5.10, second virtual electromagnetic torque value Motor torque calculation module 5.12, electromagnetic torque estimation DETAILED DESCRIPTION
[0016] The following is a clear and complete description of the specific embodiments of the present invention in conjunction with the accompanying drawings. The described embodiments are only a portion of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative work are within the scope of protection of the present invention.
[0017] Figure 1 The system composition and control block diagram of the present invention are described. The joint module includes a motor drive controller 1, a motor 2, a reducer 3, a reducer output terminal 3.1, a position sensor 4, and a control algorithm module 5 running in the main control chip of the motor drive controller 1 circuit board; the output signal of the motor drive controller 1 is output to the motor 2, and the motor current i is converted to abc 6 is output to the control algorithm module 5; the output end of the motor 2 is connected to the reducer 3; a position sensor 4 is set at the output end of the reducer 3, and the end position information θ07 is output to the control algorithm module 5; the control algorithm module 5 outputs the control pulse width PWM signal 8 to the motor drive controller 1, completing a closed-loop control system.
[0018] The joint module consists of a motor drive controller 1, such as open-source designs like oDrive and MIT Joint; a motor 2, typically a surface-mount permanent magnet motor;
[0019] The reducer 3 is generally a planetary reducer, a harmonic reducer, or an RV reducer, and includes an output terminal 3.1 of the reducer and a position sensor 4 at the output terminal. The main control chip of the motor drive controller 1 circuit board is generally an ARM 32-bit controller, which runs a control algorithm 5; wherein the control algorithm 5 collects system feedback information, including the motor current i abc 6, which is generally collected by a sampling resistor or current sensor via the ADC pin of the single-chip microcomputer; including the end position θ07 collected by the position sensor 4, which is generally collected by position sensing technologies such as optical encoder disks and magnetic encoders via the orthogonal ABZ pins, SPI pins or ADC pins of the single-chip microcomputer. The control algorithm is calculated and run to provide control information to the system, including the control pulse width PWM signal 8 required by the three-phase driver; the control algorithm 5 includes a position sensorless observation algorithm, specifically a high-frequency injection positionless observer 5.1, which is also known in the art as the HFI algorithm, or / and a back-electromotive force positionless observer 5.2, which is also known in the art as the bemf algorithm, and a reducer ratio conversion module 5.3, which performs multiplication calculations based on the mechanical reduction ratio of the reducer, and sometimes nonlinear backlash compensation can also be considered; the control algorithm 5 also includes a virtual axis model 5.4, which is a rotational inertia model that runs virtually in the control algorithm. The specific transfer function will be described in detail below. The virtual axis is used to generate the angle information required for motor control.
[0020] Depending on the motor speed, the high-frequency injection position-less observer 5.1 operates at low speed, and the back-EMF position-less observer 5.2 operates at medium and high speed. At the same time, only one of the observers is running. The specific operation is determined by a pre-defined speed threshold, such as 100 rad / s, and the high-speed switching selection 5.5 selects one of the two to obtain the estimated position of the position-less observer. The estimated position, minus the angle information After the difference is calculated, the first virtual electromagnetic torque value is obtained through the control module K1(s)5.8
[0021] According to the end position information θ07, combined with the mechanical ratio of the reducer 3, the motor angle conversion value is obtained through the reducer ratio conversion module 5.3 The converted value, minus the angle information After the difference is calculated, the second virtual electromagnetic torque value is obtained through the control module K2(s)5.10
[0022] The input of the virtual axis model 5.4 is the sum of three parts, including the electromagnetic torque estimate calculated by the motor torque calculation module 5.12. The first virtual electromagnetic torque value The second virtual electromagnetic torque value The transfer function of the virtual axis model 5.4 is Among them J m The virtual moment of inertia of the virtual axis should be as close as possible to the real physical moment of inertia of the robot joint module. It is determined by the length of the robot arm, the ratio of the reducer, and the weight of the object grasped by the robot arm. The general value is 0.0001kg / m2. The output of the virtual axis model 5.4 is the angle information
[0023] Control module K1(s) 5.8 and control module K2(s) 5.10 may employ one of the following control methods: P control (proportional control); PI control (proportional-integral control); or PID control (proportional-integral-derivative control). Other common control methods, including sliding film control, may also be employed.
[0024] In order to make the system fully operational, the system also uses a general motor joint module control module, including a motion control module, which generally adopts PID closed-loop position control, a coordinate transformation module, a current control module, a coordinate inverse transformation module, and an SVPWM module. The above modules are well known in the art, for example, they are provided in STM's open source control code and MIT joint's open source control method; the system also includes a filtering separation module for separating the high-frequency injection component and the motor fundamental component, which can usually adopt a first-order low-pass filter; the system also includes a high-frequency injection generator module and a high-frequency injection position-free observer 5.1, which are well known in the art and can be seen in TI's C2000 series high-frequency injection example code; the system also includes a back-EMF position-free observer 5.2, which is well known in the art, such as Microchip's back-EMF observation manual and Infineon's TLE987x series microcontroller software solution that can be used for free evaluation.
Claims
1. A motor control method for a robot joint module with a single end position sensor, characterized by: The joint module comprises a motor drive controller (1), a motor (2), a reducer (3), a reducer output terminal (3.1), a position sensor (4), and a control algorithm module (5) running in a main control chip of a circuit board of the motor drive controller (1); the output signal of the motor drive controller (1) is output to the motor (2), and the motor current i abc (6) is output to the control algorithm module (5); the output end of the motor (2) is connected to the reducer (3); a position sensor (4) is set at the output end of the reducer (3), and the end position information θ0 (7) is output to the control algorithm module (5); the control algorithm module (5) outputs the control pulse width PWM signal (8) to the motor drive controller (1), completing a closed-loop control system.
2. The motor control method of the end-point single position sensor robot joint module according to claim 1, characterized in that: The control algorithm module (5) collects system feedback information, including at least the motor current i abc (6) and the terminal position information θ0 (7) collected by the position sensor (4); at the same time, the control algorithm module (5) provides control information to the motor drive controller (1), including at least the control pulse width PWM signal (8) required by the three-phase driver; the control algorithm module (5) includes a position sensorless observation algorithm, specifically a high-frequency injection positionless observer (5.1), a back-electromotive force positionless observer (5.2), a reducer ratio conversion module (5.3), a virtual axis model (5.4), and a method for generating the angle information required for motor control. (9).
3. The motor control method of the end-point single position sensor robot joint module according to claim 2, characterized in that: Depending on the motor speed, the high-frequency injection position-free observer (5.1) operates at low speed, and the back-EMF position-free observer (5.2) operates at medium and high speed. At the same time, only one of the observers is running. The specific operation status is determined by the pre-defined speed threshold, and one of the two is selected by the high-speed switching selection (5.5) to obtain the estimated position of the position-free observer. (5.6); The estimated position, minus the angle information (9) After the difference is calculated, the first virtual electromagnetic torque value is obtained through the control module K1(s) (5.8) (5.9).
4. The motor control method of the end-point single position sensor robot joint module according to claim 2, characterized in that: The terminal position information θ0 (7) is combined with the mechanical transformation ratio of the reducer (3) and the reduction ratio conversion module (5.3) to obtain the motor angle conversion value. (5.7); The converted value, minus the angle information (9) After the difference is calculated, the second virtual electromagnetic torque value is obtained through the control module K2(s) (5.10) (5.11).
5. The motor control method of the end-point single position sensor robot joint module according to claim 2, characterized in that: The input of the virtual axis model (5.4) is composed of three parts, including the electromagnetic torque estimate calculated by the motor torque calculation module (5.12) (5.13), the first virtual electromagnetic torque value (5.9), the second virtual electromagnetic torque value (5.11); the transfer function of the virtual axis model (5.4) is Among them J m is the virtual moment of inertia of the virtual axis, s is the Laplace frequency variable, and its value is as close as possible to the real physical moment of inertia of the robot joint module; the output of the virtual axis model (5.4) is the angle information (9).
6. The motor control method of the end-point single position sensor robot joint module according to claim 3, characterized in that: The control module K1(s) (5.8) adopts one of the following control methods: P control, i.e. proportional control; PI control, i.e. proportional integral control; PID control, i.e. proportional integral differential control.
7. The motor control method of the end-point single position sensor robot joint module according to claim 3, characterized in that: The control module K2(s) (5.10) adopts one of the following control methods: P control, i.e. proportional control; PI control, i.e. proportional integral control; PID control, i.e. proportional integral differential control.
8. The motor control method of the end-point single position sensor robot joint module according to claim 1, characterized in that: The control algorithm module (5) also adopts a general motor joint module control module, including a motion control module, a filter separation module, a coordinate transformation module, a current control module, a coordinate inverse transformation module, an SVPWM module, and a high-frequency injection generator module.