Control moment gyroscope resonant frequency identification and suppression control method

Abstract

The invention relates to a control moment gyroscope resonant frequency identification and suppression control method, which enables CMG software to autonomously use fast Fourier transform (FFT) to carry out resonant frequency identification online and select a wave limiter parameter to suppress the frequency, can effectively improve the fluctuation of the rotating speed of a frame motor, and has the advantages of high automation in the process and high reliability. An interaction instruction is provided for a user to improve the flexibility of the method, the frame control precision of motor control can be improved, and the attitude control of a spacecraft is improved.

Description

Technical Field

[0001] The invention relates to a control moment gyro resonance frequency identification and suppression control method, belonging to the technical field of motor control. Background Art

[0002] The control moment gyro (CMG) is an important actuator for the attitude control of spacecraft, with the advantages of large output torque and high control accuracy. CMG consists of a low-speed frame motor and a high-speed rotor motor. The low-speed frame motor uses a permanent magnet synchronous motor (PMSM) to change the direction of the angular momentum generated by the high-speed rotor motor. CMG is one of the main sources of micro-vibration on the spacecraft. In order to reduce the impact of CMG micro-vibration on the spacecraft, a flexible vibration isolation mechanism needs to be installed between the CMG and the spacecraft. While suppressing the transmission of CMG vibration interference, the vibration isolation mechanism also introduces a resonant frequency in the low-frequency range (less than 50Hz), which is several thousand times the normal operating speed of the low-speed frame motor. When this frequency is introduced into the closed-loop control system of the frame motor, the output torque of the motor will fluctuate, reducing the control accuracy and stability of the frame, and affecting the attitude control effect of the spacecraft. Summary of the invention

[0003] The technical problem solved by the present invention is: in view of the problem that the traditional control method in the current prior art easily causes the output torque fluctuation of the spacecraft, a method for identifying and suppressing the resonant frequency of the control torque gyroscope is proposed.

[0004] The present invention solves the above technical problems by the following technical solutions:

[0005] A control moment gyro resonant frequency identification and suppression control method, comprising:

[0006] The sampling flag is set by default, sampling is enabled in the moment gyro control interrupt program, and sampling is stopped when the number of sampled data reaches the standard;

[0007] The identification flag is set by default, and the sampling data is processed in the moment gyro control loop program to obtain the frequency and amplitude analysis results of the corresponding quantity;

[0008] Determine the resonant frequency based on the frequency and amplitude analysis results;

[0009] Select the notch filter parameters corresponding to the resonant frequency;

[0010] A notch filter is added according to the notch filter parameters, and resonance suppression processing is performed on the target current output by the torque gyro motor through the notch filter.

[0011] The method to set the sampling flag and perform data sampling is:

[0012] Run the torque gyro control interrupt program. When the sampling enable is received in the interrupt program, the sampling flag SampleEnable is set to True, and the identification flag is set to False by default.

[0013] When the sampling flag SampleEnable changes to True, the current speed data is saved at the preset sampling frequency;

[0014] When the number of sample enable data reaches the specified number N, the sampling flag SampleEnable is set to False and sampling is stopped.

[0015] During the sampling process, the torque gyro motor operates stably at a preset speed with a preset sampling frequency as the maximum frequency, and the frequency resolution required for the spectrum analysis of the sampled data is determined according to the specified number N and the preset sampling frequency;

[0016]

[0017] In the formula, f s is the preset sampling frequency, f r The frequency resolution required for spectrum analysis of sampled data.

[0018] The method for processing sampling data and obtaining analysis results is:

[0019] When the number of adopted data reaches the specified number N, the identification flag ParamIdentiEnable is set from the default value False to True;

[0020] In the loop program, when the identification flag ParamIdentiEnable is True, a fast Fourier transform is performed on the N sampled data to obtain the analysis results of N sets of frequencies and corresponding amplitudes;

[0021] Set the identification flag ParamIdentiEnable to False to end the sampling data processing.

[0022] When the identification flag ParamIdentiEnable is set to False, the fast Fourier transform is stopped. The Fourier transform is only executed in the loop program and is not performed in the interrupt program.

[0023] The method to determine the resonant frequency is:

[0024] According to the N groups of frequency and amplitude analysis results and the prior knowledge of resonant frequency, Find the maximum amplitude Amp and its corresponding frequency Freq in the group, and use the frequency Freq as the resonant frequency;

[0025] The prior knowledge of the resonant frequency range is historical data and empirical data.

[0026] The method for selecting the notch filter parameters is:

[0027] Pre-store N / 2 sets of notch filter parameter lookup tables Table[N / 2], at a frequency of Hz and the frequency interval is f r Within the frequency range, when an instruction to automatically select parameters according to the frequency identification result is received, a parameter FilterParam of the corresponding frequency is selected from the notch filter parameters according to the resonant frequency Freq.

[0028] If a command of a user-specified parameter is received, the command frequency data is selected from the N / 2 sets of notch filter parameter lookup tables.

[0029] The transfer function of the notch filter is:

[0030]

[0031] In the formula, ω n is the notch center frequency, d is the notch depth, and M is the amplitude of the notch center frequency. db =20lg|d|, ξ is the notch filter width parameter;

[0032] The method of automatically selecting parameters is: The result is rounded to an integer as the index idx of the parameter lookup table array, and Table[idx] is used as the parameter FilterParam corresponding to Freq;

[0033] The instruction of the user-specified parameter is implemented as the index idx of the specified lookup table array, that is, the specified frequency Freq. By using the method of automatic parameter selection, idx is calculated from Freq, and finally the parameter FilterParam is obtained.

[0034] The advantages of the present invention compared with the prior art are:

[0035] A control moment gyro resonant frequency identification and suppression control method provided by the present invention can enable CMG software to autonomously use fast Fourier transform (FFT) to perform online resonant frequency identification and select notch filter parameters to suppress the frequency, which can effectively improve the fluctuation of frame motor speed. According to the user's requirement for specifying the suppression frequency, the design can receive user instructions, and the notch filter parameters corresponding to a certain frequency can be specified by the user instructions. The whole process is highly automated and easy to operate, and interactive instructions are also provided to the user to improve the flexibility of the method. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A block diagram of the motor closed-loop control method provided for the invention;

[0037] Figure 2 Bode plot of notch filter provided for the invention;

[0038] Figure 3 A schematic diagram of the control moment gyro resonant frequency identification and suppression control execution steps provided by the invention; DETAILED DESCRIPTION

[0039] A control moment gyro resonant frequency identification and suppression control method enables the CMG software to autonomously use fast Fourier transform (FFT) to identify the resonant frequency online and select limiter parameters to suppress the frequency, which can effectively improve the fluctuation of the frame motor speed. The process is highly automated and provides users with interactive instructions to improve the flexibility of the method, which can enhance the frame control accuracy of the motor control and improve the attitude control of the spacecraft.

[0040] The control moment gyro resonant frequency identification and suppression control method has the following steps:

[0041] The sampling flag is set by default, sampling is enabled in the moment gyro control interrupt program, and sampling is stopped when the number of sampled data reaches the standard;

[0042] The identification flag is set by default, and the sampling data is processed in the moment gyro control loop program to obtain the frequency and amplitude analysis results of the corresponding quantity;

[0043] Determine the resonant frequency based on the frequency and amplitude analysis results;

[0044] Select the notch filter parameters corresponding to the resonant frequency;

[0045] A notch filter is added according to the notch filter parameters, and resonance suppression processing is performed on the target current output by the torque gyro motor through the notch filter.

[0046] The method to set the sampling flag and perform data sampling is:

[0047] Run the torque gyro control interrupt program. When the sampling enable is received in the interrupt program, the sampling flag SampleEnable is set to True, and the identification flag is set to False by default.

[0048] When the sampling flag SampleEnable changes to True, the current speed data is saved at the preset sampling frequency;

[0049] When the number of sample enable data reaches the specified number N, the sampling flag SampleEnable is set to False and sampling is stopped.

[0050] During the sampling process, the torque gyro motor operates stably at a preset speed with a preset sampling frequency as the maximum frequency, and the frequency resolution required for the spectrum analysis of the sampled data is determined according to the specified number N and the preset sampling frequency;

[0051]

[0052] In the formula, f s is the preset sampling frequency, f r The frequency resolution required for spectrum analysis of sampled data.

[0053] The method for processing sampling data and obtaining analysis results is:

[0054] When the number of adopted data reaches the specified number N, the identification flag ParamIdentiEnable is set from the default value False to True;

[0055] In the loop program, when the identification flag ParamIdentiEnable is True, Fourier transform is performed on the N sampled data to obtain N sets of frequency and corresponding amplitude analysis results;

[0056] Set the identification flag ParamIdentiEnable to False to end the sampling data processing.

[0057] When the identification flag ParamIdentiEnable is set to False, Fourier transform is stopped. Fourier transform is only executed in the loop program and is not performed in the interrupt program.

[0058] The method to determine the resonant frequency is:

[0059] According to the N groups of frequency and amplitude analysis results and the prior knowledge of resonant frequency, Find the maximum amplitude Amp and its corresponding frequency Freq in the group, and use the frequency Freq as the resonant frequency;

[0060] The prior knowledge of the resonant frequency is historical data and experience data.

[0061] The method for selecting the notch filter parameters is:

[0062] Pre-store N / 2 sets of notch filter parameter lookup tables, at a frequency of Hz and the frequency interval is f r Within the frequency range, when an instruction to automatically select parameters according to the frequency identification result is received, a parameter FilterParam of the corresponding frequency is selected from the notch filter parameters according to the resonant frequency Freq.

[0063] If a command of a user-specified parameter is received, the command frequency data is selected from the N / 2 sets of notch filter parameter lookup tables.

[0064] The transfer function of the notch filter is:

[0065]

[0066] In the formula, ω n is the notch center frequency, d is the notch depth, and M is the amplitude of the notch center frequency. db =20lg|d|, ξ is the notch filter width parameter.

[0067] The method of automatically selecting parameters is, The result is rounded off and taken as the index idx of the parameter lookup table array, then Table[idx] is the parameter FilterParam corresponding to Freq.

[0068] The instruction for the user to specify the parameter can be implemented as specifying the index idx of the lookup table array, or can be implemented as specifying the frequency Freq, and then using the method of claim 10 to calculate idx from Freq to finally obtain the parameter FilterParam.

[0069] The following is further described in conjunction with the accompanying drawings and preferred embodiments of the specification:

[0070] In the current embodiment, the "loop + interrupt" structure is used through the CMG control software, and the interrupt period is in the order of hundreds of microseconds. This completes the real-time tasks related to motor control, and the machine time is not enough to complete the "identification-parameter update-suppression" functions sequentially within the interrupt. The software controls the process of online identification and suppression by setting flags, and uses the fragmented time of the interrupt to complete the frequency identification in multiple times in the loop, and completes the resonance suppression algorithm in each cycle in the interrupt. First, make the low-speed frame motor operate at a speed of ω0, and sample the speed every Δt, for a total of N times. Then, perform a fast Fourier transform (FFT) on the N sampled data to obtain the frequency distribution of the speed. Given the number of sampling points N and the sampling frequency It can be seen that the frequency resolution of the FFT result is The maximum frequency is f s After that, analyze the frequency distribution of FFT and find the frequency with the largest amplitude, which is the resonant frequency f o Finally, a notch filter is used to suppress the resonant frequency. A notch filter is a band-stop filter with a very narrow stop band, which can quickly attenuate a signal of a certain frequency. In the control software, 0 to f are pre-stored at a fixed frequency interval Δf. s The notch filter parameters within the frequency range form a parameter lookup table with frequency as the index. A notch filter is added after the software speed loop. The parameters used by the notch filter are selected from the lookup table and are consistent with the resonant frequency fo The parameters corresponding to the closest frequency. In some cases, the user hopes to specify the suppression of a certain frequency. The software is designed to accept the user's instructions to specify the notch filter parameters corresponding to a certain frequency.

[0071] Improved methods for online identification and suppression of the resonant frequency of the control moment gyro, such as Figure 3 As shown, specifically:

[0072] (1) Sampling. When the software receives the sampling enable in the interrupt program, it sets the SampleEnable flag to True. The default value of this flag is False. When the software determines that SampleEnable is True, it uses f s =100Hz sampling frequency to save N = 1024 speed data, where N is generally selected as an integer power of 2. When the number of sampled data reaches N, set SampleEnable to False to stop sampling. During the sampling process, the PMSM operates at a speed of ω0 = 1° / s. f s Determines the maximum frequency that can be obtained by spectrum analysis. In this application, 100Hz is selected; f s and N together determine the frequency resolution f that can be obtained by spectrum analysis. r The relationship between the three is The frequency resolution in this application is approximately 0.1 Hz.

[0073] (2) Spectrum analysis. When the number of samples reaches N, set the parameter identification flag ParamIdentiEnable to True. The default value of this flag is False. When the software determines that ParamIdentiEnable is True in the loop program, it performs a fast Fourier transform (FFT) on the N sampled data to obtain N groups of frequency and corresponding amplitude analysis results. In this application, the data is 0 to 100 Hz with an interval of 0.1 Hz. After completing the FFT, set ParamIdentiEnable to False, and the software will no longer repeat the FFT. The formula for FFT is: in The number of multiplications required is Compared with the discrete Fourier transform N 2 The larger N is, the more times the number of multiplications is reduced, which means more computing time can be saved. The structure of the software is "loop + interrupt", and the interrupt cycle is T cyc= 400us, complete current sampling, speed sampling, motor control, instruction processing and other tasks with high real-time requirements, and process other tasks with low real-time requirements in the loop. Since FFT calculation cannot be completed in one interrupt, in order to make full use of the remaining fragmented time of the interrupt and not affect the functions in the interrupt, FFT is performed in the loop, and the start and end process of FFT can be controlled by using the ParamIdentiEnable flag.

[0074] (3) Determine the resonant frequency. Analyze the N groups of results obtained by FFT. In order to eliminate the DC component in the signal and avoid introducing unnecessary noise, the software combines the prior knowledge of the resonant frequency and calculates the resonant frequency from the following equation: Find the maximum amplitude value Amp and its corresponding frequency Freq in the group, which is the resonant frequency. Using the prior knowledge of the resonant frequency to narrow the analysis scope can not only save time but also avoid introducing the uncertainty of irrelevant frequencies.

[0075] (4) Select the notch filter parameters. The software pre-stores N / 2 sets of notch filter parameter lookup tables, corresponding to The frequency range is Hz, and the frequency interval is f r When receiving an instruction to automatically select parameters based on the frequency identification result, the software selects the parameter FilterParam of the corresponding frequency from the notch filter parameters according to the resonant frequency Freq determined in the previous step. If receiving an instruction to specify parameters by the user, the software selects from the notch filter parameters according to the frequency given by the instruction.

[0076] (5) Resonance suppression. Figure 1 As shown in the figure, a notch filter is added between the speed loop and the current loop of the software so that the target current output by the speed converter does not contain the resonant frequency component. The user can set whether to introduce the notch filter through the command. The transfer function of the notch filter is: where ω n is the notch center frequency, d is the notch depth, and represents the amplitude of the notch center frequency M db =20lg|d|, ξ is related to the half bandwidth. The amplitude-frequency characteristics and phase-frequency characteristics of the Bode diagram are as follows Figure 2 As shown, the notch frequency is 25Hz. The discretization of G(s) uses bilinear transformation

[0077] Although the present invention has been disclosed as above in the form of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall fall within the protection scope of the technical solution of the present invention.

[0078] The contents not described in detail in the specification of the present invention belong to the common knowledge of those skilled in the art.

Claims

1. A method for identifying and suppressing the resonant frequency of a control moment gyro, characterized in that include: The sampling flag is set by default, sampling is enabled in the moment gyro control interrupt program, and sampling is stopped when the number of sampled data reaches the standard; The identification flag is set by default, and the sampling data is processed in the moment gyro control loop program to obtain the frequency and amplitude analysis results of the corresponding quantity; Determine the resonant frequency based on the frequency and amplitude analysis results; Select the notch filter parameters corresponding to the resonant frequency; A notch filter is added according to the notch filter parameters, and resonance suppression processing is performed on the target current output by the torque gyro motor through the notch filter.

2. A control moment gyro resonant frequency identification and suppression control method according to claim 1, characterized in that: The method to set the sampling flag and perform data sampling is: Run the torque gyro control interrupt program. When the sampling enable is received in the interrupt program, the sampling flag SampleEnable is set to True, and the identification flag is set to False by default. When the sampling flag SampleEnable changes to True, the current speed data is saved at the preset sampling frequency; When the number of sample enable data reaches the specified number N, the sampling flag SampleEnable is set to False and sampling is stopped.

3. A control moment gyro resonant frequency identification and suppression control method according to claim 2, characterized in that: During the sampling process, the torque gyro motor operates stably at a preset speed with a preset sampling frequency as the maximum frequency, and the frequency resolution required for the spectrum analysis of the sampled data is determined according to the specified number N and the preset sampling frequency; In the formula, f s is the preset sampling frequency, f r The frequency resolution required for spectrum analysis of sampled data.

4. The control moment gyro resonant frequency identification and suppression control method according to claim 1, characterized in that: The method for processing sampling data and obtaining analysis results is: When the number of adopted data reaches the specified number N, the identification flag ParamIdentiEnable is set from the default value False to True; In the loop program, when the identification flag ParamIdentiEnable is True, a fast Fourier transform is performed on the N sampled data to obtain the analysis results of N sets of frequencies and corresponding amplitudes; Set the identification flag ParamIdentiEnable to False to end the sampling data processing.

5. A control moment gyro resonant frequency identification and suppression control method according to claim 4, characterized in that: When the identification flag ParamIdentiEnable is set to False, the fast Fourier transform is stopped. The Fourier transform is only executed in the loop program and is not performed in the interrupt program.

6. The control moment gyro resonant frequency identification and suppression control method according to claim 1, characterized in that: The method to determine the resonant frequency is: According to the N groups of frequency and amplitude analysis results and the prior knowledge of resonant frequency, Find the maximum amplitude Amp and its corresponding frequency Freq in the group, and use the frequency Freq as the resonant frequency; The prior knowledge of the resonant frequency range is historical data and empirical data.

7. The control moment gyro resonant frequency identification and suppression control method according to claim 1, characterized in that: The method for selecting the notch filter parameters is: Pre-store N / 2 sets of notch filter parameter lookup tables Table[N / 2], at a frequency of And the frequency interval is f r Within the frequency range, when an instruction to automatically select parameters according to the frequency identification result is received, a parameter FilterParam of the corresponding frequency is selected from the notch filter parameters according to the resonant frequency Freq.

8. A control moment gyro resonant frequency identification and suppression control method according to claim 7, characterized in that: If a command of a user-specified parameter is received, the command frequency data is selected from the N / 2 sets of notch filter parameter lookup tables.

9. A control moment gyro resonant frequency identification and suppression control method according to claim 8, characterized in that: The transfer function of the notch filter is: In the formula, ω n is the notch center frequency, d is the notch depth, and M is the amplitude of the notch center frequency. db =20lg|d|, ξ is the notch filter width parameter.

10. The control moment gyro resonant frequency identification and suppression control method according to claim 7, characterized in that: The method of automatically selecting parameters is: The result is rounded to an integer as the index idx of the parameter lookup table array, and Table[idx] is used as the parameter FilterParam corresponding to Freq; The instruction of the user-specified parameter is implemented as the index idx of the specified lookup table array, that is, the specified frequency Freq. By using the method of automatic parameter selection, idx is calculated from Freq, and finally the parameter FilterParam is obtained.

Images