Harmonic gear motor based on double encoders and control method thereof

Abstract

The invention relates to the technical field of motor control, in particular to a harmonic gear motor based on double encoders and a control method thereof, which comprises the following steps: processing linear Hall element signals output by a rotor shaft encoder, including high-frequency noise suppression, direct-current bias calibration, amplitude normalization and phase difference calibration, to obtain accurate orthogonal signals; a double-dq rotating coordinate system phase-locked loop algorithm is adopted, the calibrated orthogonal signals are processed, and the high-precision real-time angle of the motor rotor is obtained through calculation; based on a signal of an output shaft encoder, matching with a preset angle partition of the harmonic reducer, and identifying an absolute position sector where the current output shaft is located; the rotor angle and the sector angle of the output shaft are fused to generate a high-resolution absolute angle of the output shaft; and based on the fused angle information, a magnetic field orientation control algorithm is adopted to drive a motor, and closed-loop control is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of motor control, and in particular to a dual-encoder-based harmonic reduction motor and a control method thereof. Background Art

[0002] The widespread application of industrial automation, robotics, and high-precision equipment is placing increasingly stringent demands on the positioning accuracy, response speed, and control stability of motor drive systems. Especially in applications such as industrial robots, CNC machining equipment, medical imaging platforms, satellite attitude control, and optical platform adjustment, actuators must not only deliver high torque output but also possess low-speed, high-precision position control capabilities.

[0003] Traditional high-precision servo systems often use three-phase brushless motors combined with harmonic reducers to achieve high torque and low output speed. To achieve high-precision position control, existing systems generally rely on high-resolution rotor shaft encoders for field-oriented control and closed-loop control. Some high-end systems also implement dual feedback by adding a second encoder at the reducer output, further improving system accuracy and robustness.

[0004] However, existing high-performance servo motor systems using a dual-encoder structure are usually costly, complex, and difficult to install and debug. For example, commercially available industrial high-precision harmonic motors often use high-cost components such as photoelectric encoders and absolute magnetic encoders, which are insufficient in terms of size, price, and maintainability. At the same time, obtaining the output shaft angle often relies on installing a second encoder at the rear end of the reducer. This has a complex structure, high dependence on customized components, and lacks versatility and modularity. In addition, in low-cost magnetic encoding systems built on Hall elements, due to factors such as signal noise, offset, amplitude inconsistency, and phase error, directly using the inverse tangent method to calculate the angle often results in large errors, making it difficult to meet precision control requirements, thus limiting its application in the industrial field. Summary of the Invention

[0005] The object of the present invention is to provide a harmonic reduction motor based on dual encoders and a control method thereof, so as to solve the problems raised in the above background technology.

[0006] To achieve the above-mentioned objectives, the present invention provides the following technical solutions: a harmonic reduction motor based on a dual encoder, comprising: a housing, a motor arranged in the housing, and the motor adopts a hollow shaft structure; a harmonic reducer arranged on one side of the motor, the input end of the harmonic reducer is connected to the output shaft of the motor through a wave generator, and the output end of the harmonic reducer is also a hollow shaft structure; a return rod, the return rod passes through the hollow shaft of the motor and is connected to the output shaft of the harmonic reducer, and extends to the rear of the motor; a radial magnet is fixed to the end of the return rod; an output shaft encoder is arranged on one side of the return rod, for detecting the magnetic field signal generated by the radial magnet and obtaining the absolute angle of the output shaft; a rotor shaft encoder is arranged at the end of the hollow shaft of the motor, for obtaining the position angle of the motor rotor.

[0007] Preferably, it also includes a return cover arranged at the output end of the harmonic reducer, which is used to fix the return rod to the center position of the output shaft; and a return fixing plate arranged on the other side of the motor, which is used to support and limit the return rod through a bearing to prevent it from radial shaking.

[0008] Preferably, the rotor shaft encoder includes a plurality of Hall elements fixedly arranged around the radial magnet, and the plurality of Hall elements are evenly distributed along the circumferential direction.

[0009] Preferably, a control method for a harmonic reduction motor based on a dual encoder comprises the following steps: processing the linear Hall element signal output by the rotor shaft encoder, including high-frequency noise suppression, DC bias calibration, amplitude normalization, and phase difference calibration, to obtain an accurate orthogonal signal; using a dual DQ rotating coordinate system phase-locked loop algorithm to process the calibrated orthogonal signal and calculate the high-precision real-time angle of the motor rotor; based on the signal of the output shaft encoder, matching it with the preset angle partition of the harmonic reducer to identify the absolute position sector of the current output shaft; fusing the rotor angle with the output shaft sector angle to generate a high-resolution absolute angle of the output shaft; based on the fused angle information, using a magnetic field oriented control algorithm to drive the motor to achieve closed-loop control.

[0010] Preferably, the high-frequency noise suppression uses a low-pass filter to filter the Hall signal.

[0011] Preferably, the DC offset calibration adopts a periodic integration and averaging method, calculating the signal average value within multiple periods to remove the offset.

[0012] Preferably, the phase difference calibration uses ellipse fitting and least square method to fit the normalized signal points, thereby calculating the phase difference between the two Hall signals.

[0013] Preferably, the motor communicates with an external host computer or a main control system via a bus to transmit the absolute angle of the output shaft, the rotor angle and the motor driving state parameters in real time.

[0014] Preferably, the method includes functional modules for rotor encoder self-calibration, output shaft encoder automatic calibration, and motor pole pair number and electrical angle automatic calibration.

[0015] Beneficial effects of the present invention: 1. The present invention connects a feedback rod to the output end of the harmonic reducer and detects the magnetic field through a radial magnet at the rear end of the motor in conjunction with an output shaft encoder. This achieves the acquisition of the absolute position of the output shaft, eliminating the need to add complex structures or additional sensors at the rear end of the reducer, thereby reducing system complexity and manufacturing costs.

[0016] 2. The present invention adopts a hollow encoder structure at the rotor shaft end. By arranging four linear Hall elements with 90° intervals to detect the radial magnet magnetic field, the present invention combines filtering, normalization, ellipse fitting, phase-locked loop and other algorithms to perform high-precision signal processing, thereby achieving high-resolution angle detection at low cost and meeting the real-time accuracy requirements of magnetic field oriented control.

[0017] 3. The present invention combines the rotor angle and the output shaft sector angle to construct a dual-encoder fusion model, which can not only accurately obtain the final absolute angle of the output shaft, but also compensate for errors such as flexibility and return clearance that may exist in the harmonic reducer, significantly improving the positioning accuracy and control robustness of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 This is a schematic diagram of the overall structure of a harmonic reduction motor based on dual encoders according to an embodiment of the present invention; Figure 2 This is an exploded view of the overall structure of a harmonic reduction motor based on dual encoders according to an embodiment of the present invention; Figure 3 This is an exploded view of the overall structure of a harmonic reduction motor based on dual encoders according to an embodiment of the present invention; Figure 4 This is a waveform diagram of a DC biased Hall signal and corresponding position error of a control method for a harmonic reduction motor based on dual encoders according to an embodiment of the present invention; Figure 5This is a waveform diagram of Hall signals with unequal amplitudes and corresponding position errors in a control method for a harmonic reduction motor based on dual encoders according to an embodiment of the present invention; Figure 6 This is a waveform diagram of the error in the Hall signal of a control method for a harmonic reduction motor based on dual encoders according to an embodiment of the present invention; Figure 7 This is a waveform diagram of high-frequency noise after passing through a low-pass filter in a control method for a harmonic reduction motor based on dual encoders according to an embodiment of the present invention; Figure 8 This is a waveform diagram after DC bias calibration of a control method for a harmonic reduction motor based on dual encoders according to an embodiment of the present invention.

[0020] Figure 9 This is a waveform diagram after signal amplitude calibration of a control method for a harmonic reduction motor based on dual encoders according to an embodiment of the present invention.

[0021] Figure 10 This is a phase difference waveform diagram calculated by a control method for a harmonic reduction motor based on dual encoders in an embodiment of the present invention.

[0022] Figure 11 This is a waveform diagram of the results of a complete algorithm of a control method for a harmonic reduction motor based on dual encoders in an embodiment of the present invention.

[0023] The following are marked in the figure: housing (100), motor (101), harmonic reducer (102), feedback rod (103), radial magnet (104), output shaft encoder (105), rotor shaft encoder (106), feedback cover (107), feedback fixing plate (108), Hall element (109). DETAILED DESCRIPTION

[0024] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to specific embodiments.

[0025] It should be noted that, unless otherwise defined, the technical or scientific terms used in the present invention should have the usual meanings understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0026] See also Figures 1 to 3 The present invention provides a technical solution for a harmonic reduction motor based on a dual encoder: a housing 100 is used to provide overall mounting support and external protection for a motor 101 and a reduction mechanism; a three-phase brushless motor 101 is arranged in the housing 100, and the rotor shaft of the motor 101 adopts a hollow shaft structure to facilitate the penetration arrangement of a subsequent signal return structure; a harmonic reducer 102 is installed on one side of the output end of the motor 101, and the input end of the harmonic reducer 102 is connected to the output shaft of the motor 101 through a wave generator, and the output end of the harmonic reducer 102 is also a hollow shaft structure for allowing the return structure to pass through; a return rod 103, one end of which is fixedly connected to the center of the output shaft of the harmonic reducer 102, penetrating The harmonic reducer 102 is connected to the hollow shaft of the motor 101 and extends axially to the rear position of the motor 101; the radial magnet 104 is installed at the tail end of the return rod 103, and is used to provide a stable radial magnetic field that can be sensed by the magnetic encoder; the output shaft encoder 105 is arranged at the tail of the motor 101, located radially outside the radial magnet 104, and is used to detect the magnetic field change signal to calculate the absolute angle information of the output shaft of the harmonic reducer 102; the rotor shaft encoder 106 is arranged at the end position of the hollow shaft of the motor 101, which includes four linear Hall elements 109 arranged in a ring, which is used to detect the change of the magnetic field of the magnet as the rotor rotates, so as to obtain the angular position of the rotor of the motor 101 in real time. Among them, the feedback rod 103 is connected to the center of the output shaft through the feedback cover 107 at the output end of the harmonic reducer 102, and the radial magnet 104 and the output shaft encoder 105 form a non-contact magnetic induction cooperation to achieve high-robustness detection of the output shaft angle; the rotor shaft encoder 106 and the drive board are integrated and arranged inside the motor 101, and can obtain high-precision rotor angles through signal sampling and calibration algorithms, providing a feedback basis for magnetic field oriented control.

[0027] A feedback cover 107 is mounted on the output end of the harmonic reducer 102. This cover 107 has a central positioning hole coaxially aligned with the output shaft of the harmonic reducer 102. One end of the feedback rod 103 passes through this hole and is rigidly connected to the output shaft. This structure precisely secures the feedback rod 103 to the center of the reducer's output shaft, ensuring axial transmission of the angle signal and improving the stability and repeatability of output shaft angle detection. A feedback fixing plate 108 is also installed on the other side of the motor 101, away from the harmonic reducer 102. This fixing plate is screwed to the rear of the motor 101 and has a through-hole for accommodating and supporting the end of the feedback rod 103. A rolling bearing or sliding bearing assembly is incorporated into the feedback fixing plate 108 to radially support and axially limit the feedback rod 103, preventing it from shaking due to centrifugal force or structural looseness during operation of the motor 101. This ensures stable magnetic field transmission and improves the sampling accuracy of the angle signal.

[0028] See also Figures 4 to 11 The present invention provides a dual-encoder-based harmonic reduction motor control method, comprising the following steps: processing the linear Hall element signal output by the rotor shaft encoder, including high-frequency noise suppression, DC offset calibration, amplitude normalization, and phase difference calibration, to obtain an accurate orthogonal signal; using a dual-DQ rotating coordinate system phase-locked loop algorithm to process the calibrated orthogonal signal and calculate the high-precision real-time angle of the motor rotor;

[0029] The output shaft encoder signal is matched with the harmonic reducer's preset angle partitions to identify the current absolute position sector of the output shaft. The rotor angle and the output shaft sector angle are then fused to generate a high-resolution absolute output shaft angle. Based on this fused angle information, a field-oriented control algorithm is used to drive the motor, achieving closed-loop control. In this embodiment, to construct a low-cost, high-resolution motor rotor angle detection system, a hollow encoder structure based on magnetic induction is placed at the end of the motor rotor shaft.

[0030] This encoder utilizes a radially magnetized magnet, fixed to the end of the rotor and rotating synchronously with the motor's rotor. The magnet's magnetic field is distributed radially, and four linear Hall elements are arranged circumferentially and evenly spaced below it. These Hall elements are spatially spaced at 90-degree intervals, resulting in the magnetic flux density changes sensed during the magnet's rotation forming sinusoidal signals with a 90-degree phase shift.

[0031] Specifically, according to the physical arrangement order of the Hall elements (for example, in a clockwise direction), the outputs of the four Hall elements are recorded as A, B, A - 、B -, where A and B are a set of mutually orthogonal signals, A - and B - is another set of orthogonal signals. Assuming A = sin(θ) and B = cos(θ), the inverse tangent function can be used to calculate the current rotor angle θ as shown below: Ideally, the four Hall elements should output waveform signals with equal amplitude, equal frequency and strict phase difference of 90 degrees, so any set of orthogonal signals (such as A, B or A - 、B - ) can be used for angle calculation.

[0032] This method achieves high-precision angle acquisition through spatial sensing of pure analog signals. It has a simple structure and fast response, avoiding the high requirements of traditional photoelectric encoders for axis alignment and high-precision cleaning. It is a low-cost position feedback solution that is very suitable for hollow motors.

[0033] In this embodiment, the preliminary calculation of the rotor angle can be achieved by using the inverse tangent function, as shown in the aforementioned formula: θ = arctan(A / B) Among them, A and B are the orthogonal signals output by the two Hall elements (109), and ideally should correspond to sin(θ) and cos(θ) respectively.

[0034] However, directly using the inverse tangent function to calculate the angle is highly dependent on the ideal characteristics of the Hall effect signal. Any errors or interference introduced into A and B, such as high-frequency noise, DC offset, amplitude inconsistency, or phase deviation, will significantly affect the accuracy and stability of the final angle calculation.

[0035] First, consider the influence of high-frequency noise. Assume that the measured A and B signals are in the form of ideal signals superimposed with noise: A = sin(θ) + n a B = cos(θ) + n β Among them, n a 、n β A high-frequency noise signal with zero mean is superimposed on the orthogonal signal channel. At this point, after calculating the angle using the inverse tangent function, the resulting angle signal θ will produce nonlinear noise propagation, and its covariance σ²θ can be estimated as follows: σ²θ = cos²(θ) · σ² a + sin²(θ) · σ² β This formula shows that even if the original A and B channel noises are small, they may be amplified after the inverse tangent nonlinear function transformation, resulting in jitter in the angle solution result and decreased accuracy.

[0036] If there is a DC offset in the signal, a fundamental frequency (1 times the electrical frequency) sinusoidal error will appear in the inverse tangent angle: Among them, factors such as zero drift of the linear Hall sensor, zero drift of the operational amplifier, and inaccurate power supply of the operational amplifier or A-D conversion circuit will all cause a DC offset in the final Hall signal sampling result. In this case, the position calculation result and its error can be expressed as: Where, is the electrical angle calculated based on the measured signal; 、 are the DC components of the two signals respectively; is the measurement error of the electrical angle, such as Figure 4 shown.

[0037] Secondly, if the amplitudes of the two Hall signals are unequal (for example, due to differences in sensitivity or installation height), a double fundamental frequency (twice the electrical frequency) error will be introduced, causing the calculated angle to fluctuate twice with the rotor position. Factors such as the varying sensitivity of the linear Hall sensors, inconsistent amplification factors of the operational amplifier circuits, and different installation heights of the Hall sensors can all contribute to unequal amplitudes of the two sampled Hall signals. In this case, the position calculation result and its error are: Where, is the ratio of the amplitudes of the two Hall signals.

[0038] Inconsistent amplitudes of the two Hall signals will produce an error of twice the fundamental frequency in the calculated position, as shown in Figure 5.

[0039] Third, when the installation angle of the Hall element does not strictly reach 90° orthogonality, there will be a DC component and a double fundamental frequency error in the calculated angle: It is difficult to ensure that the interval between the two Hall sensors is exactly 90 degrees when installing them. This will cause the fundamental phase of the Hall signal to be non-orthogonal. In this case, the Hall signal can be expressed as: 90 degrees plus is the phase difference between the two Hall signals. By performing arc tangent processing on these two signals, the corresponding position can be obtained.

[0040] From the formula, we can see that the position error can be expressed as:

[0041] From the above analysis, it can be seen that the non-orthogonal phase of the fundamental wave of the two Hall signals will cause the calculated rotor position to have DC error and double fundamental frequency error. The corresponding waveform is as follows: Figure 6 shown.

[0042] In summary, if the Hall effect signals are not preprocessed, these errors will cause periodic fluctuations or constant offsets in the angle calculation results, affecting control accuracy and system stability.

[0043] Therefore, the encoder control logic is divided into two independent processing stages: calibration and phase-locked loop processing. The calibration stage is responsible for systematic error compensation of the raw Hall effect signals, including DC offset removal, amplitude normalization, and phase difference calibration, providing a high-quality quadrature signal foundation for the subsequent phase-locked loop angle tracking algorithm.

[0044] During the calibration process, the control system drives the motor to rotate continuously at a low speed (about 1-10 rps) and stably collects the four-way Hall element signals (A, B, A - 、B - The entire calibration process lasts approximately 50 seconds, during which no real-time response is required, allowing the use of a stronger low-pass filter to suppress high-frequency noise. In this embodiment, four low-pass filters with a cutoff frequency of 50 Hz are used to filter the four Hall signals. Figure 7 , is the result of inputting a signal with high-frequency noise and passing it through a low-pass filter.

[0045] The first thing to calibrate is the DC bias. The DC bias only needs to calculate the average value of the signal within one cycle, but considering that the motor speed during calibration is not necessarily completely uniform, the average value of each cycle within 10s is calculated and the average value of the average value of all cycles is finally taken to be the DC bias of the signal. To calculate the average value of each cycle, it is only necessary to record the zero crossing point of the signal (the zero crossing point here is 1.65V) and calculate the area-time ratio between any three zero crossing points. The following is the output result of the correction after the signal with DC bias is input and the 10s DC bias calculation is performed, as shown below. Figure 8 .

[0046] The second calibration is the signal amplitude. The goal here is to normalize the signal to a signal with a DC bias of 1.65 and an amplitude of 1.65. The core of amplitude calibration is to extract the maximum and minimum values ​​of the signal. Since the DC bias has been aligned in the first step, this step only requires extracting the maximum value. The maximum value of the signal is extracted by the selection method, that is, if the input signal is greater than the current maximum value recorded, the current input is recorded as the maximum value for 10 seconds. The maximum value recorded at this time can be considered as the amplitude-mean of the signal. Then calibrate the signal to an amplitude of 1.65 by geometric scaling. The result after correction: Figure 9 .

[0047] So far, the input signal has been normalized, and the purpose of these normalization steps is to serve the third step of phase calibration. The invention uses an ellipse fitting method to calculate the phase difference between two signals that should be orthogonal.

[0048] Assume the two signals are: Expanding y and substituting x into it gives:

[0049] After squaring, we get: The general equation of an ellipse is: By simultaneously acquiring normalized signals, we can obtain sufficient data points covering multiple cycles to reduce random errors. These points can then be fitted into an ellipse, and the least squares method can be used to derive the basic parameters of the ellipse, which can be obtained as cos(φ) and sin(φ). Fitting the ellipse to a general parametric equation: The phase difference φ can be calculated by the following formula: Using this method to calculate once every 1 second, the phase difference obtained will fluctuate, but it will be close to the actual value: Figure 10 As shown, A more accurate phase difference can be obtained through digital average filtering.

[0050] After normalizing the two signals A and B and obtaining their phase difference, an orthogonal signal representing the angle θ can be obtained.

[0051] Then α and β are orthogonal. Similarly, A- and B- can also produce a pair of orthogonal signals α- and β-.

[0052] After obtaining the orthogonal signal, the inverse tangent can theoretically be used to directly obtain θ. However, if the encoder is slightly moved, the amplitude or DC bias will change, which will cause the obtained θ to be distorted. However, using the positive and negative sequence dual dq rotating coordinate system phase-locked loop DDSRF-PLL method to obtain θ, as long as the previously calculated phase difference does not change, that is, the Hall element (109) soldered on the PCB does not move relative to each other, a relatively correct θ can be obtained.

[0053] DDSRF-PLL is an improved phase-locked loop technology designed for unbalanced grid conditions. Its core concept is to separate the positive and negative sequence components of the grid voltage using dual synchronous rotating coordinate systems (positive and negative sequence), and eliminate the coupling interference between the two using a decoupling network, thereby achieving accurate tracking of grid frequency and phase.

[0054] The input signal is the orthogonal signals α and β obtained by Clark transformation of the original three-phase signal ABC with a phase difference of 120 degrees. We can take α and β as input. The following is the result of the whole algorithm. Figure 11 shown.

[0055] The blue image shows the actual angle signal, the yellow image shows the angle signal derived using the inverse tangent, and the orange image shows the signal derived using a phase-locked loop (PLL). Although the PI phase of the PLL introduces some hysteresis, it is within an acceptable range and does not require recalibration if only the amplitude and offset of the input signal change.

[0056] Overall system: signal selection: whether it is too low-pass - DC bias calibration - amplitude calibration - ellipse fitting to calculate phase difference - calculation of orthogonal α and β - phase-locked loop to obtain angle.

[0057] As for the output shaft encoder, in this invention, the output shaft encoder does not need too high precision, it only needs to be able to distinguish 101 output shaft angle sectors (the reduction ratio of the harmonic reducer is 1:101). Therefore, in this invention, the output shaft encoder uses a finished magnetic encoder chip, and obtains the approximate output shaft angle by detecting the radial magnet returned by the output shaft. This angle is then compared with the 101 pre-calibrated sector angle dividing lines to obtain the sector where the current output shaft angle is located, and finally the rotor shaft encoder angle is integrated to obtain a higher-precision output shaft angle.

[0058] The motor drive algorithm uses FOC control (field oriented control). In addition to the conventional FOC motor drive function, it adds hollow encoder self-calibration, output shaft encoder (105) automatic calibration, motor pole pair and electrical angle calibration, output shaft angle fusion, custom CAN communication and other functions.

[0059] Working principle: 1. Power drive part The motor is a three-phase brushless DC (BLDC) motor driven by field-oriented control (FOC). Three-phase orthogonal current flows through the stator windings, generating a rotating magnetic field that interacts with the permanent magnet rotor to achieve continuous and smooth electromagnetic torque output. The rotor utilizes a hollow shaft structure, and its power output shaft is connected to a wave generator, driving a harmonic reducer.

[0060] 2. Deceleration output part The motor's output shaft is connected to the harmonic reducer's wave generator, which drives the flexspline to elastically deform within the rigid wheel, creating rolling engagement. Due to the small tooth difference between the flexspline and the rigid wheel, a high reduction ratio (e.g., 1:101) is achieved per rotation, resulting in lower output speeds and higher position resolution while maintaining torque. The harmonic reducer's output shaft also features a hollow structure to facilitate angular feedback.

[0061] 3. Angle detection and control part The present invention adopts a dual encoder system to construct a full closed-loop control structure, which are: Rotor shaft encoder: Installed at the end of the motor's hollow shaft, it uses multiple linear Hall elements to detect the radial magnetic field of the magnet rotating with the rotor. After filtering, normalization, and ellipse fitting calibration, it calculates the motor's electrical angle with high precision for position closed-loop feedback in the FOC control algorithm. Output shaft encoder: It is connected to the output end of the harmonic reducer through a thin rod that passes through the return line. A radial magnet is set at the tail end, and its angular position is detected by a magnetic induction absolute encoder to identify the actual physical angle of the output shaft after deceleration. The data obtained by the two encoders are integrated through an algorithm to form a dual-channel high-precision feedback mechanism with "electrical angle + mechanical angle". This not only realizes stable drive control inside the motor, but also achieves high-resolution positioning and compensation of the output shaft angle, effectively solving problems such as nonlinear errors and return clearance caused by the harmonic reduction mechanism.

[0062] 4. System closed-loop control The controller receives external commands via CAN or other buses, calculates the difference between the target angles, and generates the control target. Based on the feedback signal, it employs a nested control strategy combining field-oriented control (FOC), current loop, velocity loop, and angle loop to finely adjust the three-phase PWM current. Ultimately, the output shaft accurately responds to control commands, achieving high-precision and robust position servo control.

[0063] It should be understood by those skilled in the art that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present invention (including the claims) is limited to these examples. Within the scope of the present invention, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the present invention as described above, which are not provided in detail for the sake of simplicity.

[0064] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A harmonic reduction motor based on dual encoders, characterized in that: include: Housing (100), A motor (101) is disposed in the housing (100), wherein the motor (101) adopts a hollow shaft structure; A harmonic reducer (102) is provided on one side of the motor (101), wherein the input end of the harmonic reducer (102) is connected to the output shaft of the motor (101) via a wave generator, and the output end of the harmonic reducer (102) is also a hollow shaft structure; A thin return rod (103), the thin return rod (103) passes through the hollow shaft of the motor (101), is connected to the output shaft of the harmonic reducer (102), and extends to the rear of the motor (101); A radial magnet (104) is fixed to the end of the return rod (103); An output shaft encoder (105) is provided on one side of the feedback rod (103) and is used to detect the magnetic field signal generated by the radial magnet (104) to obtain the absolute angle of the output shaft; A rotor shaft encoder (106) is provided at the end of the hollow shaft of the motor (101) and is used to obtain the position angle of the rotor of the motor (101).

2. The dual-encoder harmonic reduction motor according to claim 1, characterized in that: It also includes a return cover (107) provided at the output end of the harmonic reducer (102) and used for fixing the return thin rod (103) to the center position of the output shaft; The return fixing plate (108) provided on the other side of the motor (101) is used to support and limit the return thin rod (103) through a bearing to prevent it from shaking in the radial direction.

3. The harmonic reduction motor based on dual encoders according to claim 2 is characterized in that: The rotor shaft encoder (106) comprises a plurality of Hall elements (109) fixedly arranged around the radial magnet (104), wherein the plurality of Hall elements (109) are evenly distributed along the circumferential direction.

4. A dual-encoder-based harmonic reduction motor control method, characterized in that: The dual-encoder-based harmonic reduction motor according to any one of claims 1 to 3 comprises the following steps: Process the linear Hall element signal output by the rotor shaft encoder, including high-frequency noise suppression, DC offset calibration, amplitude normalization, and phase difference calibration, to obtain accurate orthogonal signals; A dual-DQ rotating coordinate system phase-locked loop algorithm is used to process the calibrated orthogonal signals and calculate the high-precision real-time angle of the motor rotor. Based on the signal of the output shaft encoder, it is matched with the preset angle partition of the harmonic reducer to identify the absolute position sector of the current output shaft; Fuse the rotor angle with the output shaft sector angle to generate a high-resolution output shaft absolute angle; Based on the fused angle information, a magnetic field oriented control algorithm is used to drive the motor to achieve closed-loop control.

5. The dual-encoder-based harmonic reduction motor control method according to claim 4, characterized in that: The high-frequency noise suppression uses a low-pass filter to filter the Hall signal.

6. The dual-encoder-based harmonic reduction motor control method according to claim 5, characterized in that: The DC offset calibration adopts a periodic integration and averaging method, which calculates the signal average value within multiple periods to remove the offset.

7. The dual-encoder-based harmonic reduction motor control method according to claim 6, characterized in that: The phase difference calibration uses ellipse fitting and least square method to fit the normalized signal points, thereby calculating the phase difference between the two Hall signals.

8. The dual-encoder-based harmonic reduction motor control method according to claim 4, characterized in that: The motor communicates with an external host computer or a main control system via a bus to transmit the absolute angle of the output shaft, the rotor angle and the motor driving state parameters in real time.

9. The dual-encoder-based harmonic reduction motor control method according to claim 4, characterized in that: The method includes the functions of rotor encoder self-calibration, output shaft encoder automatic calibration, and motor pole pair number and electrical angle automatic calibration.

Images