A method of flywheel dual redundant phase commutation control and a flywheel system control system

Abstract

The application relates to a reaction flywheel double-redundancy commutation control method and a reaction flywheel system control system, relates to the field of aviation, effectively overcomes the technical bottleneck of "Hall element failure, i.e. shutdown" in the existing flywheel system, and realizes double-redundancy commutation control of the Hall sensor and the incremental encoder only by software calibration without increasing hardware modification and testing. If the Hall sensor normally operates, the reaction flywheel is controlled according to the Hall sensor commutation logic; if the Hall sensor fails, it is further judged whether the reaction flywheel is restarted or not, if not, the reaction flywheel is controlled according to the incremental encoder commutation logic; if yes, a zero search operation is performed until the zero search ends, the zero search operation is exited, and the reaction flywheel is controlled according to the incremental encoder commutation logic. The application is suitable for processing of the reaction flywheel which is a core actuator of a satellite attitude control system.

Description

Technical Field

[0001] This invention relates to the field of aviation, and more specifically to the field of satellite reaction flywheel control technology. Background Technology

[0002] The reaction flywheel is the core actuator of the satellite attitude control system. It generates reaction torque through the change of its own angular momentum to achieve maneuvering and stability control of the satellite attitude. Its performance directly affects key mission indicators such as the accuracy of satellite Earth observation and the pointing stability of communication antennas.

[0003] In reaction-flywheel control systems, the commonly used architecture employs a brushless DC motor as the power output unit, coupled with an optical encoder as the primary speed feedback element. This combination effectively improves the flywheel's speed control accuracy. The Hall sensor, besides serving as a backup for speed feedback, is also the sole source of motor commutation information. Therefore, this design carries a single point of failure risk, making it difficult to meet the application requirements of high-reliability satellites. If the Hall sensor fails due to space radiation effects, mechanical vibration, or electrical aging, the flywheel will stop rotating, leading to satellite attitude control failure, triggering the entire satellite to enter a safety mode, and potentially causing mission failure.

[0004] The existing solutions to the above problems are as follows: First, a hardware redundancy approach is adopted, which improves system reliability by adding backup Hall sensors. However, this method is limited by the layout space of the flywheel control circuit board and the overall size of the flywheel, and it also leads to a significant increase in development costs.

[0005] Second, the Hall sensor-free solution, also known as sensorless control, cannot currently meet the stringent requirements of high-precision satellite control.

[0006] In summary, existing reaction flywheel control technologies suffer from the risk of single-point failure of Hall sensors, and their mainstream hardware redundancy solutions are also significantly limited by high cost and space constraints, making it difficult to meet the design requirements of high reliability, miniaturization, and low cost for reaction flywheels. Summary of the Invention

[0007] This invention overcomes the technical bottleneck of "system failure due to Hall element failure" in existing flywheel systems. This invention provides the following solution: Option 1: A dual-redundant commutation control method for a reaction flywheel, characterized in that... If the Hall sensor is operating normally, the reaction flywheel is controlled according to the Hall sensor commutation logic; If the Hall sensor fails, the next step is to determine whether to restart the reaction flywheel. If not restarted, the reaction flywheel is controlled according to the commutation logic of the incremental encoder; If restarting, perform the following steps: Perform zero-finding operation until the zero-finding is completed, and then exit the zero-finding operation. Control the reaction flywheel according to the commutation logic of the incremental encoder.

[0008] Furthermore, in one embodiment of the present invention, the incremental encoder commutation logic refers to: If Q ’ If ≤Q1, then the reaction flywheel operates according to the conduction sequence of the Hall states corresponding to Q1 when rotating forward in the corresponding relationship table, and operates according to the conduction sequence of the Hall states corresponding to Q1 when rotating in reverse in the corresponding relationship table. If Q n-1 <Q ’ ≤Q n When the reaction flywheel rotates forward, it follows the relationship table Q. n The Hall effect operates according to the corresponding sequence of Hall states during forward rotation, and according to Q during reverse rotation. n The corresponding Hall state conduction sequence operates during the reversal; If Q6 < Q ’ When the reaction flywheel rotates forward, it operates according to the conduction sequence of the Hall state corresponding to Q6 in the corresponding relationship table when rotating forward, and when rotating in reverse, it operates according to the conduction sequence of the Hall state corresponding to Q6 when rotating in reverse. Q ’ The remainder of the pulse count value in the current Hall state and the theoretical difference M between the adjacent pulse count values; Q1, Q n Q6 and Q6 are the six calibration pulse count values ​​in the corresponding table, and n takes the value of 2 to 5.

[0009] Furthermore, in one embodiment of the present invention, the redundant commutation control process of the reaction flywheel involves an incremental encoder, a Hall sensor, and a DSP processor. The DSP processor includes a capture module eCAP and an orthogonal encoder pulse module eQEP. The correspondence table is obtained through the following steps: Step S1: For each rotation of the flywheel body, the incremental encoder outputs an index pulse signal QEPI. When the index pulse signal QEPI arrives, the quadrature encoding pulse module eQEP resets its built-in position counter. Step S2: The capture module eCAP captures the edge of the Hall sensor output signal during one revolution of the flywheel body, and uses the current Hall state and the corresponding absolute pulse count value as the calibration data when the reaction flywheel rotates forward. The edges of the Hall sensor output signal include a rising edge and a falling edge; Step S3: Based on the absolute pulse count values in the calibration data during the forward rotation of the reaction wheel, obtain 6 calibration pulse count values. Sort the obtained calibration pulse count values as Q1 < Q2 < Q3 < Q4 < Q5 < Q6, and use the sorted calibration pulse count values and the corresponding Hall states as the forward rotation encoder pulse marking table. The 6 calibration pulse count values include Q1, Q2, Q3, Q4, Q5, and Q6. Step S4: In the case of the reaction wheel rotating in reverse, repeat Step S2 and Step S3 to obtain the reverse rotation encoder pulse marking table. Step S5: Based on the forward rotation encoder pulse marking table and the reverse rotation encoder pulse marking table obtained in Step S3 and Step S4, use the calibration pulse count values as indexes to correspond the data in the two encoder pulse marking tables, and obtain the forward and reverse correspondence table.

[0010] Further, in an embodiment of the present invention, in Step S3, based on the absolute pulse count values in the calibration data during the forward rotation of the reaction wheel, obtaining 6 calibration pulse count values specifically is as follows: Assume that the total number of pulses output by the flywheel body per rotation is N times, and the number of pole pairs of the reaction wheel is P; then the flywheel body experiences 6*P Hall state switches per rotation. The theoretical difference between adjacent pulse count values in the same Hall state is M = N / P; the same Hall state corresponds to P groups of the absolute pulse count values and Hall states. After taking the remainder of the absolute pulse count values in the same Hall state with respect to the theoretical difference M of adjacent absolute pulse count values, and then averaging these remainders, the calibration pulse count value corresponding to this Hall state is obtained. The remaining Hall states are processed in the same way to obtain six calibration pulse count values, which are recorded as Q1, Q2, Q3, Q4, Q5, and Q6 in ascending order after sorting.

[0011] Further, in an embodiment of the present invention, the zero seeking operation is as follows: Control the reaction wheel to rotate at an extremely low speed, and continuously capture the index pulse signal QEPI output by the encoder. If the index pulse signal QEPI is captured, perform a reset operation on the position counter of the quadrature encoder pulse module eQEP, and the zero seeking ends.

[0012] Solution 2: A control system for a reaction wheel system. The control system for the reaction wheel system includes a processor DSP, a Hall sensor signal acquisition unit, an incremental encoder signal acquisition unit, a current acquisition unit, a drive circuit, and a communication unit. The processor DSP internally embeds a data processing module, a pulse width modulation module ePWM, a capture module eCAP, a quadrature encoder pulse module eQEP, and an AD conversion module implemented by a computer program. The Hall sensor signal acquisition unit is used to acquire the signal output by the Hall sensor and send it to the capture module eCAP of the processor DSP. The incremental encoder signal acquisition unit is used to acquire the signal output by the encoder and send it to the quadrature encoder pulse module eQEP of the processor DSP. The current acquisition unit is used to acquire the bus current of the flywheel-driven brushless DC motor, amplify it, and send it to the AD conversion module of the processor DSP. The drive circuit is used to generate a DC motor drive signal based on the received PWM signal and send it to the brushless DC motor. When the computer program is executed, it performs any of the commutation methods described above.

[0013] Furthermore, in one embodiment of the present invention, the processor DSP is further embedded with a communication unit, which is used to interact with the satellite platform central machine.

[0014] The dual-redundant commutation control method for reaction flywheels described in this invention is based on Hall sensors and incremental encoders. It effectively overcomes the technical bottleneck of "failure of Hall element equals system crash" in existing flywheel systems. Furthermore, it requires no additional hardware modifications or testing; dual-redundant commutation control of the Hall sensor and incremental encoder can be achieved solely through software calibration. Specific beneficial effects include: 1. The reaction flywheel commutation control method described in this invention is a dual-redundant commutation control method. This method represents a completely different technical solution from existing control methods. The main difference lies in the fact that existing technologies rely solely on Hall sensors for reaction flywheel commutation control, which carries a single-point-of-failure risk. Adding backup Hall sensors introduces hardware redundancy issues, while not using Hall sensors currently fails to meet the requirements for high-precision satellite control. To address these issues, this invention designs a dual-redundant commutation control method for reaction flywheels. Commutation is achieved through logic that determines the winding conduction sequence based on the absolute pulse count interval. This serves as a fault response strategy for the reaction flywheel. When a Hall sensor failure is detected, this commutation control method is immediately activated to ensure continuous normal operation of the flywheel while maintaining unaffected flywheel speed control accuracy. Furthermore, this method can achieve dual-redundant commutation control of the Hall sensor and incremental encoder solely through software calibration.

[0015] 2. The incremental encoder commutation logic described in this invention serves as a fault response strategy for the reaction flywheel. It is immediately activated upon detecting a Hall sensor failure to ensure continuous normal operation of the flywheel. The reaction flywheel employs a brushless DC motor and is controlled using a six-step commutation method. The Hall sensor state changes every 60 electrical degrees of motor rotation, marking the arrival of the commutation point. Since the correspondence between the commutation point and the encoder's absolute pulse count value is predetermined through ground calibration, in subsequent Hall sensor failure modes, the system can determine the winding conduction sequence based on the current absolute pulse count range, thereby achieving precise commutation control. This method effectively ensures that the control accuracy of the flywheel speed remains unaffected.

[0016] The method described in this invention is applicable to the processing of reaction flywheels, the core actuators of satellite attitude control systems. Attached Figure Description

[0017] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is the reaction flywheel dual-redundant commutation control diagram described in Implementation Method 1.

[0018] Figures 2 to 7 Figure 2 shows a comparison of flywheel speed accuracy at different speeds based on Hall effect commutation logic and flywheel speed accuracy based on incremental encoder commutation logic, as described in Implementation Method 2. Figure (a) shows the comparison of flywheel speed accuracy during forward rotation, and Figure (b) shows the comparison of flywheel speed accuracy during reverse rotation. The flywheel speed accuracy based on the Hall effect commutation strategy is the flywheel speed accuracy based on Hall effect sensor commutation logic, while the flywheel speed accuracy based on the backup commutation strategy is the flywheel speed accuracy based on incremental encoder commutation logic. Figures 2 to 7 These correspond to speeds of ±10rpm, ±50rpm, ±500rpm, ±2000rpm, ±3000rpm, and ±4000rpm, respectively.

[0019] Figure 8 This is a schematic diagram of the encoder calibration result display interface as described in Implementation Method 3.

[0020] Figure 9 This is a schematic representation of the correspondence between the absolute pulse count value and the Hall state of the encoder described in Embodiment 3.

[0021] Figure 10 This is a schematic diagram of the encoder absolute pulse count value under different Hall states during one revolution of the motor as described in Embodiment 3.

[0022] Figure 11 This is a linear relationship diagram of the absolute pulse count value of the encoder under the same Hall state as described in Embodiment 3.

[0023] Figure 12 This is a measured diagram of the encoder calibration results described in Implementation Method 3.

[0024] Figure 13 This is a schematic diagram showing the correspondence between the absolute pulse count value and the Hall state of the measured encoder described in Embodiment 3.

[0025] Figure 14 This is a flowchart of the encoder pulse marking method based on a Hall sensor as described in Embodiment 3.

[0026] Figure 15 This is the overall block diagram of the control circuit of the reaction flywheel system described in Implementation Method 5. Detailed Implementation

[0027] Various embodiments of the present invention will now be clearly and completely described with reference to the accompanying drawings. The embodiments described with reference to the drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0028] Implementation Method 1: The reaction flywheel dual-redundant commutation control method described in this implementation method, in this implementation method, as follows... Figure 1 As shown, If the Hall sensor is operating normally, the reaction flywheel is controlled according to the Hall sensor commutation logic; If the Hall sensor fails, the next step is to determine whether to restart the reaction flywheel. If not restarted, the reaction flywheel is controlled according to the commutation logic of the incremental encoder; If restarting, perform the following steps: Perform zero-finding operation until the zero-finding is completed, and then exit the zero-finding operation. Control the reaction flywheel according to the commutation logic of the incremental encoder.

[0029] In this embodiment, since the reaction flywheel is a long-term on-orbit operating device, if the following situations occur during its on-orbit operation, the reaction flywheel restart operation needs to be performed: 1. Fault-correcting restarts triggered by abnormal flywheel conditions: including but not limited to control logic errors caused by single-event upsets, flywheel speed deviations, or communication interruptions; 2. Mission-adaptive restart triggered by satellite attitude control mode switching: When the ground control center sends a command to the satellite to adjust its attitude control mode according to the satellite's on-orbit mission requirements, the flywheel needs to be restarted to adapt to the new control strategy. 3. Safety restart triggered by satellite autonomous protection logic: If the satellite system triggers autonomous safety protection logic, such as attitude instability or exceeding the fault tolerance threshold, the system needs to be restored by restarting the flywheel.

[0030] In the above situations, restarting the reaction flywheel is used to clear the fault state, restore control functions, or ensure task continuity and system safety.

[0031] The dual-redundant commutation control method for the reaction flywheel described in this embodiment is based on the logic of determining the winding conduction sequence according to the range of absolute pulse count values. As a fault response strategy for the reaction flywheel, it is activated immediately when the Hall sensor failure is detected to ensure that the flywheel continues to work normally while maintaining the flywheel speed control accuracy.

[0032] Implementation Method Two: This implementation method further defines the reaction-driven flywheel dual-redundant commutation control method described in Implementation Method One. In this implementation method, the incremental encoder commutation logic refers to: If Q ’ If ≤Q1, the reaction flywheel operates according to the conduction sequence of the Hall states corresponding to Q1 during forward rotation when rotating forward, and according to the conduction sequence of the Hall states corresponding to Q1 during reverse rotation when rotating in reverse. If Q n-1 <Q ’ ≤Q n When the reaction flywheel rotates forward, it follows the relationship table Q. n The Hall effect operates according to the corresponding sequence of Hall states during forward rotation, and according to Q during reverse rotation. n The corresponding Hall state conduction sequence operates during the reversal; If Q6 < Q ’ When the reaction flywheel rotates forward, it operates according to the conduction sequence of the Hall state corresponding to Q6 in the corresponding relationship table when rotating forward, and when rotating in reverse, it operates according to the conduction sequence of the Hall state corresponding to 6 when rotating in reverse. Q ’ Q1, Q' is the remainder of the theoretical difference M between the pulse count value in the current Hall state and the adjacent pulse count values. n Q6 and Q6 are the six calibration pulse count values ​​in the corresponding table, and n takes the value of 2 to 5.

[0033] This embodiment provides an example where, based on the correspondence table obtained in Embodiment 3, the incremental encoder commutation logic of the reaction flywheel can be obtained, as follows: When the pulse count value is Q, first calculate the remainder: Q' = mod(Q, 3600), then judge according to the following interval and perform the corresponding commutation operation: If Q'≤467 (Q1), then it will work in the original Hall state 1 conduction sequence when rotating forward, and in the original Hall state 5 conduction sequence when rotating backward. If 467 (Q1) < Q’ ≤ 1079 (Q2), when rotating forward, it works according to the conduction sequence of the original Hall state 3, and when rotating backward, it works according to the conduction sequence of the original Hall state 1; If 1079 (Q2) < Q’ ≤ 1642 (Q3), when rotating forward, it works according to the conduction sequence of the original Hall state 2, and when rotating backward, it works according to the conduction sequence of the original Hall state 3; If 1642 (Q3) < Q’ ≤ 2273 (Q4), when rotating forward, it works according to the conduction sequence of the original Hall state 6, and when rotating backward, it works according to the conduction sequence of the original Hall state 2; If 2273 (Q4) < Q’ ≤ 2878 (Q5), when rotating forward, it works according to the conduction sequence of the original Hall state 4, and when rotating backward, it works according to the conduction sequence of the original Hall state 6; If 2878 (Q5) < Q’ ≤ 3452 (Q6), when rotating forward, it works according to the conduction sequence of the original Hall state 5, and when rotating backward, it works according to the conduction sequence of the original Hall state 4; If Q’ > 3452 (Q6), when rotating forward, it works according to the conduction sequence of the original Hall state 5, and when rotating backward, it works according to the conduction sequence of the original Hall state 4.

[0034] As Figures 2 to 7 shown, it is a comparison chart of the flywheel speed accuracy based on the Hall commutation logic and the flywheel speed accuracy based on the incremental encoder commutation logic at different speeds. Figure (a) is the comparison chart of the flywheel speed accuracy when rotating forward, and Figure (b) is the comparison chart of the flywheel speed accuracy when rotating backward. Among them, the flywheel speed accuracy of the Hall commutation strategy is the flywheel speed accuracy based on the commutation logic of the Hall sensor, and the flywheel speed accuracy of the backup commutation strategy is the flywheel speed accuracy based on the commutation logic of the incremental encoder. Figures 2 to 7 They respectively correspond to speeds of ±10 rpm, ±50 rpm, ±500 rpm, ±2000 rpm, ±3000 rpm and ±4000 rpm. It can be seen that the commutation logic of this incremental encoder can maintain the speed control accuracy without being affected.

[0035] This embodiment further defines the commutation logic of the incremental encoder, and illustrates the scheme of the commutation logic of the incremental encoder. As a fault response strategy for the reaction wheel, it is immediately enabled when a Hall sensor failure is detected to ensure the continuous normal operation of the flywheel.The reaction wheel uses a brushless DC motor and is controlled in a six-step commutation mode. Every 60° electrical angle of motor rotation, the Hall state changes once, and this change marks the arrival of the commutation point. Given that the correspondence between the commutation point and the absolute pulse count value of the encoder has been determined in advance through ground calibration, in the subsequent fault mode of Hall sensor failure, the system can determine the logic of the winding conduction sequence according to the current interval of the absolute pulse count value, so as to achieve precise commutation control. This method effectively ensures that the control accuracy of the flywheel speed is not affected.

[0036] Embodiment 3. This embodiment further defines the reaction wheel dual-redundancy commutation control method described in Embodiment 2. The reaction wheel redundancy commutation control process involves an incremental encoder, a Hall sensor, and a processor DSP. The processor DSP includes a capture module eCAP and a quadrature encoder pulse module eQEP. In this embodiment, the schematic diagram of the encoder calibration result is as Figure 8 shown, and the correspondence table between the absolute pulse count value of the encoder and the Hall state is as Figure 9 shown. The correspondence table is obtained through the following steps: Step S1, every time the flywheel body rotates one week, the incremental encoder outputs an index pulse signal QEPI, and the quadrature encoder pulse module eQEP resets the built-in position counter when the index pulse signal QEPI arrives; Step S2, the capture module eCAP captures the edges of the output signal of the Hall sensor during one positive rotation of the flywheel body, and takes the current Hall state and the corresponding absolute pulse count value as the calibration data when the flywheel rotates forward; The edges of the output signal of the Hall sensor include rising edges and falling edges; Step S3, based on the absolute pulse count values in the calibration data when the flywheel rotates forward, obtain 6 calibration pulse count values, sort the obtained calibration pulse count values as Q1 < Q2 < Q3 < Q4 < Q5 < Q6, and take the sorted calibration pulse count values and the corresponding Hall states as the forward rotation encoder pulse mark table; The 6 calibration pulse count values include Q1, Q2, Q3, Q4, Q5, Q6; Step S4, in the case of the flywheel rotating in reverse, repeat Step S2 and Step S3 to obtain the reverse rotation encoder pulse mark table; Step S5: Based on the forward encoder pulse mark table and the reverse encoder pulse mark table obtained in steps S3 and S4, the data in the two encoder pulse mark tables are matched with the calibration pulse count value as the index to obtain the correspondence table between forward and reverse rotation.

[0037] In this embodiment, step S3, which involves obtaining six calibration pulse count values ​​based on the absolute pulse count values ​​in the calibration data during forward rotation of the reaction flywheel, specifically involves: Let N be the total number of pulses output per revolution of the flywheel, and P be the number of pole pairs of the reaction flywheel. Then, the flywheel undergoes 6*P Hall state switching cycles per revolution. The theoretical difference between adjacent pulse counts under the same Hall state is M=N / P. The same Hall state corresponds to P groups of absolute pulse counts. The remainders of the absolute pulse counts under the same Hall state divided by the theoretical difference M are then averaged to obtain the calibration pulse counts corresponding to that Hall state. The remaining Hall states are processed in the same way to obtain six calibration pulse counts, which are sorted from smallest to largest and denoted as Q1, Q2, Q3, Q4, Q5, and Q6.

[0038] In this embodiment, the calibration result is preferably embedded in the commutation function to control the reaction flywheel based on the commutation logic of the incremental encoder. In this embodiment, the incremental encoder mentioned in step S1 preferably outputs two quadrature pulse signals QEPA and QEPB, and an index pulse signal QEPI used as a reference zero position.

[0039] In this embodiment, step S2 preferably ensures that the recorded encoder pulse count values ​​are all absolute positions relative to the zero position. Therefore, the marking process should be performed after the motor has completed at least one full forward rotation and the eQEP position counter has been reset by QEPI.

[0040] This embodiment provides an example that uses a 0.05Nms reaction flywheel from an on-orbit satellite. The specific steps are as follows: In step E1, for each revolution of the flywheel, the incremental encoder outputs an index pulse signal QEPI to identify the absolute starting position. When the index pulse signal QEPI arrives, the orthogonal encoding pulse module eQEP resets its built-in position counter, so that the subsequent position counter output values ​​are all absolute positions relative to the zero position signal.

[0041] Step E2: When the capture module eCAP captures the arrival of the edge (including rising edge and falling edge) of the Hall sensor output signal during the forward rotation of the flywheel, a capture interrupt is triggered. In the eCAP interrupt service function, a two-dimensional array is constructed. The first column of the array records the current Hall state, and the second column records the current absolute pulse count value of the encoder. The recorded data is used as the calibration data when the reaction flywheel rotates forward.

[0042] Since the designed 0.05Nms reaction flywheel has 8 pole pairs, the motor will experience 48 Hall state switching cycles per revolution. Therefore, the two-dimensional array is set to 48 rows and 2 columns.

[0043] After sending a positive speed command (e.g., +10 rpm) to the flywheel, in order to ensure the absoluteness of the recorded pulse count value relative to the zero position signal, the marking process must be carried out after the motor has completed at least one full forward rotation and the eQEP position counter has been reset by the index signal QEPI. The marking process ends when the two-dimensional array is full of data.

[0044] Step E3: The designed 0.05Nms reaction flywheel motor outputs 28800 pulses per revolution, with 8 pole pairs. Therefore, the theoretical difference between adjacent pulse counts under the same Hall state is M = 28800 / 8 = 3600, and the theoretical difference between pulse counts under adjacent Hall states is 3600 / 6 = 600. In the two-dimensional array, the same Hall state corresponds to 8 rows of data. After taking the remainder of the encoder pulse count value under the same Hall state divided by 3600, the average of these remainders is calculated to obtain the absolute pulse count value corresponding to that Hall state. The other Hall states are processed in the same way. Finally, the obtained values ​​are sorted from smallest to largest to obtain six calibration pulse count values: Q1 = 467, Q2 = 1079, Q3 = 1642, Q4 = 2273, Q5 = 2878, and Q6 = 3452.

[0045] like Figure 10 The image shows a set of Hall effect state count values ​​and encoder absolute pulse count values ​​obtained after one revolution of the motor. After processing the data, as shown... Figure 11 As shown, it can be observed that the absolute pulse count values ​​of the encoder under the same Hall state exhibit a linear relationship, consistent with the theoretical derivation. The actual calibration data processing results are as follows: Figure 12 As shown, the difference in the calibration pulse count values ​​between two adjacent Hall states is approximately 600, which is consistent with the theoretical calculation results.

[0046] Step E4: When the reaction flywheel is in reverse, repeat steps E2 and E3 to obtain the reverse encoder pulse mark table; Step E5: Based on the forward encoder pulse mark table and reverse encoder pulse mark table obtained in steps E3 and E4, the Hall state count values ​​are mapped according to the same calibration pulse count value to obtain a measured correspondence table, as shown in the table below. Figure 13 As shown; This embodiment is an encoder pulse marking method based on a Hall sensor. It obtains backup commutation information through an incremental encoder, which is key to realizing a dual redundancy mechanism between the Hall sensor and the encoder in commutation and speed measurement. This embodiment is as follows... Figure 14 As shown, by processing the index pulse signal QEPI, marking the encoder pulse based on the Hall sensor, and processing the calibration data, the calibration results of the reaction flywheel in forward and reverse rotation are obtained, and a correspondence table is obtained based on the calibration results.

[0047] This embodiment further defines the dual-redundant commutation control method for the reaction flywheel. Through the processing of the index pulse signal (QEPI) in step S1, the QEPI is used to identify the absolute starting position. When this signal arrives, the eQEP module resets its built-in position counter, ensuring that subsequent position counter output values ​​are all relative to the absolute position of this zero-position signal. Through encoder pulse marking based on the Hall sensor in step S2 and calibration data processing in step S3, a correspondence table of forward and reverse rotation calibration results for the reaction flywheel is obtained. This correspondence table forms the basis of the incremental encoder commutation logic.

[0048] Implementation Method Four: This implementation method further defines the reaction-driven flywheel dual-redundant commutation control method described in Implementation Method One. In this implementation method, the zero-finding operation is as follows: The reaction flywheel is controlled to rotate at an extremely low speed to capture the index pulse signal QEPI output by the encoder in real time. If the index pulse signal QEPI is captured, the position counter of the quadrature encoding pulse module eQEP is reset, and the zero-finding ends.

[0049] In this embodiment, the zero-finding process is executed only once during restart and does not need to be repeated until the power is cut off again.

[0050] In this embodiment, to achieve the initial start-up and zero-search rotation of the reaction flywheel under conditions of no Hall signal (Hall element failure), it can be achieved in the following way: the motor windings are sequentially turned on according to a preset forward or reverse phase conversion sequence. Through this orderly winding conduction method, the flywheel obtains a continuous starting torque and begins to rotate, thereby completing the search process for the index pulse signal.

[0051] This embodiment further defines the dual-redundant commutation control method for reaction flywheels and provides an example of the zero-finding operation scheme. If the flywheel restarts after the Hall sensor fails, in order to ensure the absoluteness of the position count value relative to the zero position signal and thus ensure accurate commutation, a zero-finding operation must be performed first to locate the index pulse.

[0052] Implementation Method 5: A reaction flywheel system control system as described in this implementation method. In this implementation method, as follows... Figure 15 As shown, The reaction flywheel system control system includes a processor DSP, a Hall sensor signal acquisition unit, an incremental encoder signal acquisition unit, a current acquisition unit, a drive circuit, and a communication unit; The processor DSP has an embedded data processing module, a pulse width modulation module (ePWM), a capture module (eCAP), a quadrature encoder pulse module (eQEP), and an AD conversion module implemented by a computer program. The Hall sensor signal acquisition unit is used to acquire the signal output by the Hall sensor and send it to the capture module eCAP of the processor DSP. The incremental encoder signal acquisition unit is used to acquire the signal output by the encoder and send it to the quadrature encoder pulse module eQEP of the processor DSP. The current acquisition unit is used to acquire the bus current of the flywheel-driven brushless DC motor, amplify it, and send it to the AD conversion module of the processor DSP. The drive circuit is used to generate a DC motor drive signal based on the received PWM signal and send it to the brushless DC motor. When the computer program is executed, the commutation method described in any one of embodiments one to four is performed.

[0053] In this embodiment, the processor DSP may also embed a communication unit, which is used to interact with the satellite platform central machine.

[0054] The communication unit is capable of general SCI / CAN communication.

[0055] In this embodiment, the driving circuit can be implemented using a drive bridge to convert the PWM signal into a DC motor drive signal.

[0056] In this embodiment, the current acquisition unit can acquire the current signal using a sampling resistor, and then amplify it through a signal filtering and amplification circuit before sending it to the AD conversion module of the processor DSP.

[0057] The preferred embodiment of this method uses a 0.05 Nms reaction flywheel on-orbit satellite.

[0058] In this embodiment, the incremental encoder is used to output an index pulse signal QEPI and two quadrature pulse signals for each revolution of the code disk.

[0059] In this embodiment, the Hall sensor is used to provide motor rotor position information.

[0060] The reaction flywheel system control board described in this embodiment can meet the control requirements of high-performance motors. By using a DSP processor as the core controller, it possesses highly integrated peripheral resources that are perfectly suited to system requirements, significantly simplifying the design of peripheral circuits. The reaction flywheel system control board in this embodiment integrates a signal acquisition and processing module, a motor drive circuit, a communication interface circuit, a power supply module, and a debugging and downloading circuit. It achieves high-precision real-time closed-loop control of the flywheel motor's speed and winding current, and simultaneously conducts bidirectional data interaction with the satellite platform central unit through a standard communication interface (RS422 / CAN), including receiving telemetry commands, reporting remote control status, and configuring system parameters.

Claims

1. A dual-redundancy commutation control method for a reaction flywheel, characterized in that: If the Hall sensor operates normally, the reaction flywheel is controlled according to the Hall sensor commutation logic; If the Hall sensor fails, it is further determined whether to restart the reaction flywheel. If not restarted, the reaction flywheel is controlled according to the incremental encoder commutation logic; If restarted, the following steps are executed: Execute a zero-finding operation until the zero-finding ends and the zero-finding operation is exited, and control the reaction flywheel according to the incremental encoder commutation logic; The incremental encoder commutation logic refers to: If Q ’ If ≤Q1, then the reaction flywheel operates according to the conduction sequence of the Hall states corresponding to Q1 when rotating forward in the corresponding relationship table, and operates according to the conduction sequence of the Hall states corresponding to Q1 when rotating in reverse in the corresponding relationship table. If Q n-1 <Q ’ ≤Q n When the reaction flywheel rotates forward, it follows the relationship table Q. n The Hall effect operates according to the corresponding sequence of Hall states during forward rotation, and according to Q during reverse rotation. n The corresponding Hall state conduction sequence operates during the reversal; If Q6 < Q ’ When the reaction flywheel rotates forward, it operates according to the conduction sequence of the Hall state corresponding to Q6 in the corresponding relationship table when rotating forward, and when rotating in reverse, it operates according to the conduction sequence of the Hall state corresponding to Q6 when rotating in reverse. Q ’ The remainder of the pulse count value in the current Hall state and the theoretical difference M between the adjacent pulse count values; Q1, Q n Q6 and Q6 are the six calibration pulse count values ​​in the corresponding table, and n takes the value of 2 to 5.

2. The reaction flywheel dual-redundant commutation control method according to claim 1, wherein the reaction flywheel redundant commutation control process involves an incremental encoder, a Hall sensor, and a DSP processor, the DSP processor including a capture module eCAP and an orthogonal encoder pulse module eQEP, characterized in that, The corresponding relationship table is obtained through the following steps: Step S1, when the flywheel body rotates one week, the incremental encoder outputs an index pulse signal QEPI, and the orthogonal encoding pulse module eQEP resets the built-in position counter when the index pulse signal QEPI arrives; Step S2, the capture module eCAP captures the edges of the output signal of the Hall sensor during one forward rotation of the flywheel body, and takes the current Hall state and the corresponding absolute pulse count value as the calibration data when the reaction flywheel rotates forward; The edges of the output signal of the Hall sensor include rising edges and falling edges; Step S3, based on the absolute pulse count values in the calibration data when the reaction flywheel rotates forward, obtain 6 calibration pulse count values, sort the obtained calibration pulse count values as Q1 < Q2 < Q3 < Q4 < Q5 < Q6, and take the sorted calibration pulse count values and the corresponding Hall states as the forward rotation encoder pulse marking table; The 6 calibration pulse count values include Q1, Q2, Q3, Q4, Q5, Q6; Step S4, in the case of the reaction flywheel rotating in reverse, repeat Step S2 and Step S3 to obtain the reverse rotation encoder pulse marking table; Step S5, based on the forward rotation encoder pulse marking table and the reverse rotation encoder pulse marking table obtained in Step S3 and Step S4, use the calibration pulse count value as the index, correspond the data in the two encoder pulse marking tables, and obtain the corresponding relationship table for forward and reverse rotations.

3. The reaction flywheel dual-redundant commutation control method according to claim 2, characterized in that, The specific method for obtaining 6 calibration pulse count values based on the absolute pulse count values in the calibration data when the reaction flywheel rotates forward in Step S3 is as follows: Assume that the total number of pulses output when the flywheel body rotates one week is N times, and the number of pole pairs of the reaction flywheel is P; then the flywheel body experiences 6*P Hall state switches when rotating one week, and the theoretical difference between adjacent pulse count values in the same Hall state is M = N / P; The same Hall state corresponds to P groups of the absolute pulse count values and Hall states. After taking the remainder of the absolute pulse count values in the same Hall state with respect to the theoretical difference M between adjacent absolute pulse count values, and then averaging these remainders, the calibration pulse count value corresponding to this Hall state is obtained. The remaining Hall states are processed in the same way to obtain six calibration pulse count values, which are sorted from small to large and denoted as Q1, Q2, Q3, Q4, Q5, Q6 respectively.

4. The reaction flywheel dual-redundant commutation control method according to claim 1, characterized in that, The zero-finding operation is: The reaction flywheel is controlled to rotate at an extremely low speed to capture the index pulse signal QEPI output by the encoder in real time. If the index pulse signal QEPI is captured, the position counter of the quadrature encoding pulse module eQEP is reset, and the zero-finding ends.

5. A control system for a reaction flywheel system, characterized in that, The reaction flywheel system control system includes a processor DSP, a Hall sensor signal acquisition unit, an incremental encoder signal acquisition unit, a current acquisition unit, a drive circuit, and a communication unit; The processor DSP has an embedded data processing module, a pulse width modulation module (ePWM), a capture module (eCAP), a quadrature encoder pulse module (eQEP), and an AD conversion module implemented by a computer program. The Hall sensor signal acquisition unit is used to acquire the signal output by the Hall sensor and send it to the capture module eCAP of the processor DSP. The incremental encoder signal acquisition unit is used to acquire the signal output by the encoder and send it to the quadrature encoder pulse module eQEP of the processor DSP. The current acquisition unit is used to acquire the bus current of the flywheel-driven brushless DC motor, amplify it, and send it to the AD conversion module of the processor DSP. The drive circuit is used to generate a DC motor drive signal based on the received PWM signal and send it to the brushless DC motor. When the computer program is executed, it performs the commutation method according to any one of claims 1-4.

6. The reaction flywheel system control system according to claim 5, characterized in that, The processor DSP also has a communication unit embedded inside, which is used to interact with the satellite platform central machine.

Images