Variable impedance vibration control method and system based on vibration dominant frequency estimation

Abstract

The application belongs to the technical field of automatic control, and particularly relates to a variable impedance vibration control method and system based on vibration main frequency estimation. The method is characterized in that a wideband excitation signal is applied and a response signal is collected, a vibration system transfer function is determined by system identification, then the transfer characteristics are analyzed, the frequency bands are divided and the impedance parameter combination is optimized to form a frequency band parameter mapping table, the vibration main frequency of the vibration system is estimated in real time by an extended Kalman filtering algorithm, the corresponding impedance parameters are determined in combination with the mapping relationship, the target control force and the target current instruction are calculated according to the real-time vibration information, the motor is driven to output the control force, and the variable impedance control is realized. The application solves the frequency domain adaptability defects of the traditional method under wideband excitation, is suitable for vehicle suspensions, precision instrument vibration isolation and other electromechanical systems that need to dynamically adjust the vibration characteristics, and can significantly optimize key indicators such as the root mean square value of the load acceleration.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of automatic control, and particularly relates to a variable impedance vibration control method and system based on vibration dominant frequency estimation, which is suitable for electromechanical systems that need to dynamically adjust vibration characteristics, such as vehicle suspensions, precision instrument vibration isolation, and other broadband vibration control systems. BACKGROUND

[0002] In the field of dynamic system control, traditional control methods face significant technical bottlenecks when dealing with frequency-varying coupling characteristics (i.e., nonlinear response caused by changes in vibration frequency over time):

[0003] (1) Existing variable impedance control mechanisms have insufficient frequency domain adaptability. Variable impedance control based on speed adjusts impedance through linear mapping, but it is difficult to adapt to nonlinear coupling effects under wideband excitation, resulting in insufficient adjustment accuracy in the sensitive frequency band; variable impedance control based on stability can ensure system stability, but its conservative constraint conditions limit the dynamic response bandwidth (i.e., the system's response ability to different frequency excitations).

[0004] (2) Traditional proportional-integral-derivative (PID) control often relies on fixed gain parameters in frequency domain performance optimization, making it difficult to adapt to dynamic changes in wideband vibration scenarios, resulting in over-damping at low frequencies and insufficient response at high frequencies; traditional optimal control methods also have similar problems, for example, linear quadratic regulator (LQR) control uses a fixed weight quadratic cost function, which results in mutual constraints among different frequency bands; H∞ control method optimizes the robustness of the system to the worst case, sacrificing the fine adjustment ability of specific frequency bands, making it difficult to achieve global optimization under wideband excitation. SUMMARY

[0005] To solve the above technical problems, the present application provides a variable impedance vibration control method and system based on vibration dominant frequency estimation, which effectively solves the contradiction of vibration transmission gain in different frequency bands of traditional vibration control systems by constructing a dynamic mapping relationship between impedance parameters and vibration dominant frequency, tracking the frequency using an extended Kalman filter algorithm, and adjusting the impedance control parameters according to the constructed mapping relationship. This method can be widely applied in the field of electromechanical system control that needs to adjust frequency-varying characteristics. Typical application scenarios include but are not limited to vehicle active suspension, precision instrument vibration isolation platform, and industrial machinery vibration suppression system.

[0006] To achieve the above purpose, the first aspect of the present application provides a variable impedance vibration control method based on vibration dominant frequency estimation, comprising the following steps:

[0007] An acceleration sensor is installed at the vibration site of the vibration system, and a displacement sensor is installed between the vibration site and the fixed base, real-time acquisition of time-domain vibration signals of the vibration system is performed, and the sampling frequency is not less than 2.5 times the highest working frequency of the system;

[0008] A vibrator is used to apply a frequency band covering the target frequency band to the vibration system. A wideband excitation signal, comprising a linear sweep frequency signal, a discrete multi-frequency synthesized signal, and a pink noise signal, is used to excite the full-frequency response of the vibration system.

[0009] The response signal of the vibration system under the action of the broadband excitation signal is collected, and the transfer function of the vibration system is extracted based on the system identification method.

[0010] Analyze the transfer function and divide the frequency bands between the frequency points where performance conflicts exist;

[0011] Based on the preset vibration control performance requirements, an impedance parameter combination that satisfies the control effect is determined for each frequency band, and a frequency band parameter mapping table is established, which represents the correspondence between frequency bands and impedance parameter combinations.

[0012] The acquired time-domain vibration signal is bandpass filtered, with the filtering range covering the effective operating frequency band of the vibration system. Then, outlier removal and moving average processing are performed on the filtered signal.

[0013] The extended Kalman filter algorithm is used to process the time-domain signal after the moving average processing to obtain the estimated vibration dominant frequency of the vibration system.

[0014] Based on the frequency band parameter mapping table, find the estimated vibration dominant frequency. Matching impedance parameters, including impedance stiffness coefficient Impedance damping coefficient and impedance inertia coefficient ;

[0015] Real-time acquisition of displacement difference of vibration system ,speed and acceleration ;

[0016] Based on the impedance parameters and the acquired displacement difference ,speed and acceleration The target control force is calculated using the following formula. :

[0017] ;

[0018] Through formula The target current command is calculated. ,in This is the torque coefficient of the motor, in units of... This represents the ratio of the motor's output torque to its input current.

[0019] The torque output of the drive motor is converted into the target control force through the transmission mechanism by the current loop PID control. It is applied to the vibration system to achieve variable impedance vibration control.

[0020] A second aspect of the present invention provides a variable impedance vibration control system based on vibration dominant frequency estimation, comprising a sensor, a processor, a driver, a motor, and a transmission mechanism; the sensor is used to collect vibration signals of the vibration system and send the collected vibration signals to the processor; the processor stores a preset frequency band parameter mapping table, and the processor processes the collected vibration signals using a signal conditioning algorithm, an extended Kalman filter algorithm, and an adaptive noise matrix adjustment algorithm to obtain an estimated value of the vibration dominant frequency of the vibration system, and determines the corresponding impedance parameters based on the estimated value of the vibration dominant frequency and the frequency band parameter mapping table, wherein the impedance parameters include an impedance stiffness coefficient. Impedance damping coefficient and impedance inertia coefficient Based on the impedance parameters and the displacement difference of the vibration system, ,speed and acceleration Information, through formulas Calculated target control force The processor controls the target force. Calculate the target current command for the motor. The driver is based on the target current command. The motor is driven to output torque, and the transmission mechanism converts the motor torque into the target control force that matches the vibration direction of the vibration system. It is applied to the vibration system to achieve real-time frequency-domain adaptive vibration control.

[0021] The beneficial effects of this invention are as follows: This invention obtains the dominant vibration frequency of the vibration system through extended Kalman filtering, and dynamically adjusts the impedance parameters by combining a preset frequency band parameter mapping table to achieve precise vibration control. This method effectively improves vibration control accuracy, resolves the gain contradictions of traditional methods in different frequency bands, and enhances system stability. Based on a frequency domain decoupling and multi-mode switching architecture, it adapts to complex working conditions under wideband excitation, exhibiting good dynamic response and stability. Furthermore, it has wide applicability and can be applied to various fields such as vehicle active suspension, precision instrument vibration isolation platforms, and industrial machinery vibration suppression, reducing vibration hazards, improving equipment performance and service life, and enhancing the comfort of the working environment. Attached Figure Description

[0022] The above and other features, advantages, and aspects of the embodiments of this disclosure will become clearer with reference to the accompanying drawings and the following detailed description. The drawings are intended to aid in understanding the present invention and are not intended to limit the scope of this disclosure. Identical or similar reference numerals represent identical or similar elements, wherein:

[0023] Figure 1 A flowchart of a variable impedance vibration control method based on vibration dominant frequency estimation;

[0024] Figure 2 A schematic diagram of a variable impedance vibration control system based on vibration dominant frequency estimation;

[0025] Figure 3 Schematic diagram of the vibration system structure;

[0026] Figure 4 Amplitude-frequency response curves of sprung mass acceleration with different impedance inertia coefficients;

[0027] Figure 5 Amplitude-frequency response curves of unsprung dynamic deformation with different impedance inertia coefficients;

[0028] Figure 6 Amplitude-frequency response curves of sprung mass acceleration with different impedance stiffness coefficients and impedance damping coefficients;

[0029] Figure 7 Amplitude-frequency response curves of unsprung dynamic deformation with different impedance stiffness coefficients and impedance damping coefficients;

[0030] Figure 8 The root mean square value of the acceleration of the sprung mass;

[0031] Figure 9 The amplitude-frequency response curve of the sprung mass acceleration. Detailed Implementation

[0032] To enable those skilled in the art to better understand the technical solutions and advantages of the present invention, the present application will be described in detail below with reference to the accompanying drawings, but this is not intended to limit the scope of protection of the present invention.

[0033] like Figure 1 As shown in the flowchart, this embodiment provides a variable impedance vibration control method based on vibration dominant frequency estimation, including the following steps:

[0034] An acceleration sensor is installed at the vibrating part of the vibration system, and a displacement sensor is installed between the vibrating part and the fixed base to collect the time-domain vibration signal of the vibration system in real time. The sampling frequency is not less than 2.5 times the highest operating frequency of the system; in some embodiments, such as Figure 3The structure shown is as follows: The vibration system includes an upper structure 4, a lower structure 5, and a fixed base 11; the upper structure 4 is the upper vibration component of the system, and the lower structure 5 is the lower vibration component; an acceleration sensor 2 is installed on the upper structure 4 to collect the acceleration signal of the upper structure 4 in real time; a displacement sensor 1 is disposed between the upper structure 4 and the lower structure 5 to measure the relative displacement between the upper structure 4 and the lower structure 5; a displacement sensor 3 is disposed between the lower structure 5 and the fixed base 11 to measure the relative displacement between the lower structure 5 and the fixed base 11.

[0035] A vibrator is used to apply a frequency band covering the target frequency band to the vibration system. A wideband excitation signal, comprising a linear sweep frequency signal, a discrete multi-frequency synthesized signal, and a pink noise signal, is used to excite the full-frequency response of the vibration system; see also Figure 3 The vibrator 10 is installed between the lower structure 5 and the fixed base 11;

[0036] The response signal of the vibration system under the action of the broadband excitation signal is collected, and the transfer function of the vibration system is extracted based on the system identification method. In some embodiments, the response signal includes the acceleration signal of the upper structure 4, the relative displacement signal between the upper structure 4 and the lower structure 5, and the relative displacement signal between the lower structure 5 and the fixed base 11, respectively collected by the acceleration sensor 2, the displacement sensor 1, and the displacement sensor 3. For ease of description, the acceleration signal of the upper structure 4 is called the sprung mass acceleration, the relative displacement between the upper structure 4 and the lower structure 5 is called the sprung dynamic deformation, and the relative displacement between the lower structure 5 and the fixed base 11 is called the unsprung dynamic deformation. The transfer function includes the transfer function of the sprung mass acceleration, the sprung dynamic deformation, and the unsprung dynamic deformation relative to the broadband excitation signal, and its expression is:

[0037] ;

[0038] ;

[0039] ;

[0040] ;

[0041]

[0042]

[0043] in, , and respectively the sprung mass acceleration, the dynamic deformation above the spring and the dynamic deformation below the spring, relative to the wideband excitation signal, denotes a complex frequency domain variable, denotes a common denominator of the three transfer functions, denotes a transfer function of the force between the superstructure 4 and the substructure 5, denotes a transfer function of the force between the substructure 5 and the fixed base 11, denotes the mass of the superstructure 4, denotes the mass of the substructure 5, denotes the stiffness coefficient of the spring 6, denotes the damping coefficient of the damper 7, denotes the stiffness coefficient of the spring 9, as Figure 3 shown, the spring 6 and the spring 9 are connected between the superstructure 4 and the substructure 5, and between the substructure 5 and the fixed base 11, respectively, for providing passive elastic support; the damper 7 is arranged between the superstructure 4 and the substructure 5, for providing passive damping action; denotes the Laplace transform of the control force of the active control actuator 8, which is installed between the superstructure 4 and the substructure 5, for applying active control force to the vibration system, denotes the displacement of the superstructure 4, denotes the displacement of the substructure 5;

[0044] The control force of the active control actuator 8 is calculated by an impedance control algorithm, the expression of which is related to the selected impedance parameters, and the Laplace transform expression of which is:

[0045] ;

[0046] wherein, denotes the stiffness coefficient in the impedance parameters, denotes the damping coefficient in the impedance parameters, denotes the inertia coefficient in the impedance parameters, denotes a complex frequency domain variable; for ease of expression, is called impedance stiffness coefficient, and is called impedance damping coefficient, and is called impedance inertia coefficient;

[0047] The transfer functions are analyzed, and frequency bands are divided between frequency points where performance conflicts exist; in some embodiments, different impedance stiffness coefficients , impedance damping coefficients and the impedance inertia coefficient The responses of different vibration systems can be obtained and analyzed.

[0048] like Figure 4 The figure shows the amplitude-frequency response curves of the spring mass acceleration for different impedance inertia coefficients. Figure 5 The figure shows the amplitude-frequency characteristic curves of unsprung dynamic deformation for different impedance inertia coefficients. The arrows in the figure indicate an increase in the impedance inertia coefficient. It can be seen that in the frequency range below 1 Hz, the impedance inertia coefficient has little effect on the acceleration of the sprung mass; in the frequency range of 1-8 Hz, the larger the impedance inertia coefficient, the smaller the amplitude of the sprung mass acceleration; while in the frequency range of 8-12 Hz, the trend is reversed, with the smaller the impedance inertia coefficient and the smaller the amplitude of the sprung mass acceleration; in the frequency range above 12 Hz, the impedance inertia coefficient has little effect on the amplitude of unsprung dynamic deformation; in the frequency range of 1-4 Hz, the larger the impedance inertia coefficient, the smaller the amplitude of the unsprung dynamic deformation; in the frequency range of 4-14 Hz, the larger the impedance inertia coefficient, the larger the amplitude of the unsprung dynamic deformation; while in the frequency range above 14 Hz, the effect is relatively small.

[0049] like Figure 6 The figure shows the amplitude-frequency response curves of sprung mass acceleration for different impedance stiffness coefficients and impedance damping coefficients. The arrows in the figure indicate an increase in the impedance stiffness coefficient. It can be seen that as the impedance stiffness coefficient increases, the first-order resonance of the sprung mass acceleration mainly varies around 1 Hz; in the frequency range of 3-8 Hz, both the minimum amplitude and its corresponding frequency increase with the increase of the impedance stiffness coefficient; the second-order resonance frequency increases with the increase of the impedance stiffness coefficient, while the resonance amplitude decreases; in the frequency range below 1 Hz, the larger the impedance damping coefficient, the smaller the amplitude; in the frequency range of approximately 2-10 Hz, the smaller the impedance damping coefficient, the smaller the amplitude; near the second-order resonance frequency, the impedance damping coefficient is negatively correlated with the amplitude; in the frequency range above 12 Hz, the smaller the impedance damping coefficient, the smaller the amplitude.

[0050] Figure 7 The figure shows the amplitude-frequency response curves of unsprung dynamic deformation for different impedance stiffness coefficients and impedance damping coefficients. The arrows in the figure indicate an increase in the impedance stiffness coefficient. It can be seen that as the impedance stiffness coefficient increases, both the first-order resonant frequency and amplitude of the unsprung dynamic deformation increase, with the main change occurring around 1 Hz. In the frequency range of 2-4 Hz, the minimum amplitude and its corresponding frequency increase with the increase of the impedance stiffness coefficient. In the frequency range above 9 Hz, the impedance stiffness coefficient has little effect. In the frequency range below 2 Hz, the larger the impedance damping coefficient, the smaller the amplitude. In the frequency range of approximately 2-8 Hz, the smaller the impedance damping coefficient, the smaller the amplitude. Near the second-order resonant frequency, the larger the impedance damping coefficient, the smaller the amplitude.

[0051] In combination with the above analysis, the frequency band is divided into three segments: 0-4Hz, 4-15Hz and 15-25Hz;

[0052] In some embodiments, according to a preset vibration control performance requirement, an impedance parameter combination satisfying a control effect is determined for each frequency segment, a frequency segment parameter mapping table is established, and the mapping table represents a corresponding relationship between the frequency segment and the impedance parameter combination; in order to suppress vibration of the vibration system and ensure stability of the vibration system, the preset vibration control performance requirement is that the sprung mass acceleration amplitude is smaller and the unsprung dynamic deformation amplitude is smaller;

[0053] The selection strategy of the impedance parameter combination is: in the 0-4Hz frequency segment, the amplitudes of the sprung mass acceleration and the unsprung dynamic deformation are suppressed simultaneously, a larger impedance inertia coefficient, a smaller impedance stiffness coefficient and a medium impedance damping coefficient are selected; in the 4-15Hz frequency segment, the amplitude of the sprung mass acceleration is preferentially suppressed, a smaller impedance inertia coefficient, a medium impedance stiffness coefficient and a smaller impedance damping coefficient are selected; in the 15-25Hz frequency segment, the amplitude of the unsprung dynamic deformation is preferentially suppressed, a larger impedance inertia coefficient, a larger impedance stiffness coefficient and a smaller impedance damping coefficient are selected. The specific frequency segment parameter mapping table is shown in Table 1:

[0054] Table Frequency segment parameter mapping table

[0055]

[0056] The collected time domain vibration signal is subjected to band-pass filtering, the filtering range covers the effective working frequency band of the vibration system, and then the filtered signal is subjected to outlier rejection and sliding average processing; in some embodiments, the band-pass filter is a Chebyshev type band-pass filter; the window length used in the sliding average processing is 50 steps;

[0057] The extended Kalman filtering algorithm is used to process the time domain signal after the sliding average processing, and the vibration main frequency of the vibration system is obtained; the extended Kalman filtering algorithm includes a state space model, a state update rule of the extended Kalman filtering algorithm loop iteration and a measurement noise variance matrix adaptively adjusted according to a signal signal-to-noise ratio; in some embodiments, the state space model includes a state vector, a state process function, a measurement variable and a measurement function; the state vector is used to describe a state of the state space model at a certain time, and its expression is:

[0058] ;

[0059] wherein, represents time, This represents the duration of one step. This represents the amplitude of the vibration process at the current moment. This represents the amplitude of the vibration process at the previous step time. The frequency of the vibration process at the current moment is called the dominant frequency. This represents the frequency of the vibration process at the previous step time. This represents the phase angle of the vibration process at the current moment. This represents the phase angle of the vibration process at the previous step time. This represents the state vector at the current moment. Represents the state vector The One portion, , This represents the sinusoidal component of the vibration displacement at the current moment. This represents the sinusoidal component of the vibration displacement at the previous step time. This represents the cosine component of the vibration displacement at the current moment. This represents the cosine component of the vibration displacement at the previous step time. This indicates the dominant vibration frequency at the current moment;

[0060] The state process function is used to describe the evolution of the model state. Specifically, based on the current state vector of the model, the state vector at the next step size can be calculated using the state process function, and its expression is:

[0061] ;

[0062] in, The state process function represents the state process function. This represents the state vector at the next step.

[0063] The measured variable The measured value acquired by the sensor is defined as follows:

[0064] ;

[0065] The measurement function Used to establish the measurement variables With the state vector The relationship between them can be expressed as:

[0066] ;

[0067] in, The measurement function can be rewritten in matrix form as follows:

[0068] ;

[0069] in, This represents the measurement matrix, with dimensions of . , ;

[0070] In some embodiments, the state update rules for the iterative extended Kalman filter algorithm include a state transition matrix update rule, a state vector estimate update rule, an estimation error covariance matrix update rule, and a Kalman gain matrix update rule.

[0071] The state transition matrix update rule is used to calculate the state transition matrix at each time step. Used to linearize state process functions It is the estimated value of the state vector. The Jacobian matrix is ​​expressed as:

[0072] ;

[0073] in, Represents the state transition matrix. This represents the estimated value of the state vector at the current moment. The state vector estimate represents the value of the state vector. The One portion, ;

[0074] The state vector estimate update rule is used to calculate the state vector estimate at each time step, and its expression is as follows:

[0075] ;

[0076] in, This represents the estimated state vector value at the previous step. This represents the measurement matrix, with dimensions of . , This represents the state process function described at the previous step time, whose input is the estimated state vector value described at the previous step time. , Let Kalman gain matrix be the Kalman gain matrix at the previous step, with dimension 1. , This represents the measurement value at the current moment;

[0077] The estimation error covariance matrix update rule is used to calculate the estimation error covariance matrix at each time step, and its expression is:

[0078] ;

[0079] in, This represents the identity matrix, with dimensions of . , denotes the Kalman gain matrix at the current time step, denotes the measurement matrix, denotes the estimation error covariance matrix at the current time step, , denotes the estimation error covariance matrix at the next time step, denotes the transpose of the state transition matrix, denotes the process noise variance matrix at the current time step, ;

[0080] The Kalman gain matrix update rule is used to calculate the Kalman gain matrix at each time step, and the expression is:

[0081] ;

[0082] wherein, denotes the Kalman gain matrix at the next time step, denotes the estimation error covariance matrix at the next time step, denotes the measurement matrix, denotes the transpose of the measurement matrix, denotes the measurement noise variance matrix at the current time step, ;

[0083] According to the calculated state vector estimation value the vibration main frequency estimation value is obtained , and the expression is:

[0084] ;

[0085] wherein, denotes the vibration main frequency estimation value at the current time step, denotes the 5th component of the state vector estimation value ;

[0086] The measurement noise variance matrix is adaptively adjusted according to the signal signal-to-noise ratio In some embodiments, the vibration main frequency estimation error is calculated, the estimation error of the measurement variable is calculated, the signal signal-to-noise ratio is calculated, and the measurement noise variance matrix is adjusted ;

[0087] The expression of the vibration main frequency estimation error is:

[0088] ;

[0089] wherein, represents the error of the vibration main frequency estimation, which is the difference between the vibration main frequency estimation before and after, represents the time length of one step, represents the vibration main frequency estimation value at the current time, represents the vibration main frequency estimation value at the last step time;

[0090] The expression for calculating the estimation error of the measurement variable is:

[0091] ;

[0092] wherein, represents the estimation error of the measurement variable, represents the first component of the state vector estimation value , which represents the sine component of the vibration displacement at the current time, represents the measurement value at the current time;

[0093] Based on the above two errors, the signal-to-noise ratio of the signal is calculated, and the expression is:

[0094] ;

[0095] wherein, represents the signal-to-noise ratio at the current time, and the numerator in the above formula represents the cumulative sum of the square of the estimation error of the measurement variable, and the denominator represents the cumulative sum of the square of the main frequency estimation error, which reflects the relative strength between the measurement noise and the system process noise;

[0096] Let the process noise variance matrix be a known constant matrix, then the measurement noise variance matrix at the current time is adaptively adjusted according to the following formula:

[0097]

[0098] wherein, represents the measurement noise variance matrix at the current time, represents the process noise variance matrix, represents the signal-to-noise ratio at the current time;

[0099] Based on the frequency band parameter mapping table, the impedance parameter matched with the vibration main frequency estimation value is searched, and the impedance parameter includes an impedance stiffness coefficient , an impedance damping coefficient and an impedance inertia coefficient ;

[0100] The displacement difference , velocity and acceleration ; in some embodiments, the displacement difference is the relative displacement between the upper structure 4 and the lower structure 5, the velocity is the relative velocity between the upper structure 4 and the lower structure 5, the acceleration is the relative acceleration between the upper structure 4 and the lower structure 5;

[0101] based on the impedance parameters and the collected displacement difference , velocity and acceleration , the target control force is calculated by the following formula:

[0102] ;

[0103] The target current command is calculated by the formula , wherein is the torque coefficient of the motor, with the unit of , representing the ratio of the motor output torque to the input current;

[0104] The current loop PID control drives the motor output torque, which is converted into the target control force by the transmission mechanism and applied to the vibration system to realize variable impedance vibration control.

[0105] As shown in Figure 2 , the present application provides a variable impedance vibration control system based on vibration main frequency estimation, comprising a sensor, a processor, a driver, a motor and a transmission mechanism; the sensor is used to collect the vibration signal of the vibration system and send the collected vibration signal to the processor; the processor has a pre-set frequency band parameter mapping table stored therein, the processor uses signal conditioning algorithm, extended Kalman filtering algorithm and adaptive noise matrix adjustment algorithm to process the collected vibration signal to obtain the vibration main frequency estimation value of the vibration system, determines the corresponding impedance parameters according to the vibration main frequency estimation value and the frequency band parameter mapping table, the impedance parameters include impedance stiffness coefficient , impedance damping coefficient and impedance inertia coefficient , and according to the impedance parameters and the displacement difference , velocity and acceleration information of the vibration system, the target control force is calculated by the formula , the processor calculates the target current command of the motor according to the target control force , and the driver calculates the target current command The motor is driven to output torque, and the transmission mechanism converts the motor torque into the target control force that matches the vibration direction of the vibration system. It is applied to the vibration system to achieve real-time frequency-domain adaptive vibration control.

[0106] In some embodiments, the sensor includes an acceleration sensor and a displacement sensor, wherein the acceleration sensor is used to acquire the acceleration signal of the vibration system, and the displacement sensor is used to acquire the displacement signal of the vibration system.

[0107] In some embodiments, such as Figure 2 The control method block diagram shown includes a processor comprising a signal conditioning module, a frequency domain estimation module, an impedance mapping module, a force generation module, and a current generation module. The signal conditioning module receives signals from the accelerometer and displacement sensor, performs filtering and smoothing operations on them, and outputs the conditioned sensor signals. The frequency domain estimation module outputs a real-time estimated vibration frequency. The impedance mapping module outputs impedance parameters according to a preset frequency band parameter mapping table; the force generation module calculates the target control force based on the selected impedance parameters. And output, the current generation module is used to receive the target control force. Calculate the target control current command .

[0108] In some embodiments, the frequency domain estimation module includes an extended Kalman filter module and an adaptive noise matrix adjustment module. The adaptive noise matrix adjustment module receives the estimated state from the previous time step and the measurement signal from the current time step, and outputs the estimated noise matrix for the current time step. The extended Kalman filter module receives the measurement signal from the current time step and the estimated noise matrix, and outputs the estimated vibration dominant frequency for the current time step. .

[0109] like Figure 8 As shown, under the input of frequency sweep sinusoidal displacement disturbance, the root mean square value of the sprung mass acceleration of the variable impedance vibration control method in this embodiment is reduced by 28.2% compared with passive vibration reduction, which is better than the traditional LQR control method. Figure 9 The amplitude-frequency response curve (solid line) of the sprung mass acceleration is further shown, indicating that the amplitude of the sprung mass acceleration remains at a low level within the 0-25Hz operating frequency band, verifying the effectiveness of real-time frequency identification and impedance parameter switching.

[0110] The above embodiments are merely one implementation method of the present invention, and not all of them. They should not be used to limit the scope of protection of the present invention. All equivalent changes or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A variable impedance vibration control method based on vibration dominant frequency estimation, characterized by, The method comprises the following steps: An acceleration sensor is installed at a vibration site of a vibration system, and a displacement sensor is installed between the vibration site and a fixed base, so as to collect time-domain vibration signals of the vibration system in real time, and a sampling frequency is not less than 2.5 times of a highest working frequency of the system; A shaker is used to apply a wideband excitation signal covering a target frequency band to a vibration system The wideband excitation signal includes a linear sweep signal, a discrete multi-tone signal, and a pink noise signal to excite a full frequency band response of the vibration system. A response signal of the vibration system under the action of the wideband excitation signal is collected, and a transfer function of the vibration system is extracted based on a system identification method; The transfer function is analyzed, and frequency bands are divided between frequency points with performance conflicts; Impedance parameter combinations meeting control effects are determined for each divided frequency band according to preset vibration control performance requirements, a frequency band parameter mapping table is established, and the frequency band parameter mapping table represents a corresponding relationship between frequency bands and impedance parameter combinations; The collected time-domain vibration signals are subjected to band-pass filtering, and a filtering range covers an effective working frequency band of the vibration system, and then the filtered signals are subjected to outlier rejection and sliding average processing; An extended Kalman filtering algorithm is used to process the time-domain signals after the sliding average processing, so as to obtain vibration main frequencies of the vibration system; Based on the frequency band parameter mapping table, the vibration main frequency estimation value is searched The matched impedance parameters include impedance stiffness coefficient , impedance damping coefficient And impedance inertia coefficient ; Real-time acquisition of displacement difference of a vibrating system , velocity , and acceleration ; Based on the impedance parameter and the displacement difference collected , velocity , and acceleration , the target control force is calculated by the following formula : ； The target current command I* is calculated by the formula wherein is the torque coefficient of the motor, with units of , representing the ratio of the motor output torque to the input current;​ The driving motor outputs torque through current loop PID control, and the torque is converted into the target control force through a transmission mechanism and applied to the vibration system to realize variable impedance vibration control.

2. The variable impedance vibration control method based on vibration dominant frequency estimation according to claim 1, wherein The state space model comprises a state vector, a state process function, a measurement variable and a measurement function; The frequency band parameter mapping table shows the impedance stiffness coefficient corresponding to the low-frequency band 0-4Hz. 15000 N / m, impedance damping coefficient 1200 Ns / m, impedance inertia coefficient It weighs 50 kg; the impedance stiffness coefficient corresponding to the mid-frequency range of 4-15 Hz. 18000 N / m, impedance damping coefficient 800 Ns / m, impedance inertia coefficient It weighs 30kg; the impedance stiffness coefficient corresponding to the high-frequency range of 15-25Hz. 22000 N / m, impedance damping coefficient 1500 Ns / m, impedance inertia coefficient It weighs 70kg.

3. The variable impedance vibration control method based on vibration dominant frequency estimation according to claim 1, wherein, The extended Kalman filtering algorithm is used to realize real-time estimation of the vibration main frequency, including a state space model, a state updating rule of a loop iteration of the extended Kalman filtering algorithm, and a measurement noise variance matrix which is adaptively adjusted according to a signal signal-to-noise ratio .

4. The variable impedance vibration control method based on vibration dominant frequency estimation according to claim 3, wherein, The state vector is used to describe a state of the state space model at a certain time, and an expression thereof is as follows: The state update rule of the extended Kalman filtering algorithm comprises a state transition matrix update rule, a state vector estimation value update rule, an estimation error covariance matrix update rule and a Kalman gain matrix update rule; ； wherein denotes the time, denotes the duration of one step, denotes the amplitude of the vibration process at the current time instant, denotes the amplitude of the vibration process at the previous time instant, denotes the frequency of the vibration process at the current time instant, i.e. the vibration dominant frequency, denotes the frequency of the vibration process at the previous time instant, denotes the phase angle of the vibration process at the current time instant, denotes the phase angle of the vibration process at the previous time instant, denotes the state vector at the current time instant, denotes the state vector at the previous time instant, denotes the k-th component of the state vector , denotes the sine component of the vibration displacement at the current time instant, denotes the sine component of the vibration displacement at the previous time instant, denotes the cosine component of the vibration displacement at the current time instant, denotes the cosine component of the vibration displacement at the previous time instant, denotes the vibration dominant frequency at the current time instant; the state process function for describing the evolution process of the model state, in particular for predicting the state vector at the next time step from the current state vector of the model at the current time step by means of the state process function, which is expressed as ； wherein, denotes the state process function, denotes the state vector at the next step time instant; the measured variable represents a measured value acquired by the sensor, defined as: ； the measurement function for establishing the measurement variable a relationship between the state vector whose expression is ； wherein denotes the measurement function, rewritten in matrix form: ； wherein denotes a measurement matrix, which has dimensions , .

5. The variable impedance vibration control method based on vibration dominant frequency estimation according to claim 3, wherein, The state vector estimation value update rule is used to calculate a state vector estimation value at each time, and an expression thereof is as follows: The state transition matrix updating rule is used to calculate the state transition matrix at each time for linearizing the state process function is the Jacobian matrix of the state vector estimate , expressed as: ； wherein, denotes a state transition matrix, denotes a state process function, denotes a current time state vector estimate, denotes the state vector estimate of the i-th component, of the i-th component, , denotes time, denotes a duration of one step; The estimation error covariance matrix update rule is used to calculate an estimation error covariance matrix at each time, and an expression thereof is as follows: ； wherein, represents a state vector estimation value at a previous step time, represents a measurement matrix, which has a dimension of , represents a state process function at a previous step time, which has an input of a state vector estimation value at a previous step time , represents a Kalman gain matrix at a previous step time, which has a dimension of , represents a measurement value at a current time; The Kalman gain matrix update rule is used to calculate a Kalman gain matrix at each time, and an expression thereof is as follows: ； wherein denotes an identity matrix of dimension , denotes the Kalman gain matrix at the current time instant, denotes the measurement matrix, denotes the estimation error covariance matrix at the current time instant of dimension , denotes the estimation error covariance matrix at the next step time instant, denotes the transpose of the state transition matrix, denotes the process noise variance matrix at the current time instant of dimension ; An expression for calculating an estimation error of the measurement variable is as follows: ； wherein, denotes the Kalman gain matrix at the next step time instant, denotes the estimation error covariance matrix at the next step time instant, denotes the measurement matrix, denotes the transpose of the measurement matrix, denotes the measurement noise variance matrix at the current time instant, of dimension ; According to the calculated state vector estimate Obtaining a vibration frequency estimate , expressed as: ； wherein represents the vibration main frequency estimation value at the current time point, represents the 5th component of the state vector estimation value at the current time point.

6. The variable impedance vibration control method based on vibration dominant frequency estimation according to claim 3, wherein, Adaptively adjusting the measurement noise variance matrix according to the signal signal-to-noise ratio , specifically comprising calculating a vibration main frequency estimation error, calculating an estimation error of the measurement variable, calculating a signal signal-to-noise ratio, and adjusting the measurement noise variance matrix ; Computing an error of vibration frequency estimation The expression for the error of vibration frequency estimation is ； wherein, represents an error of the vibration main frequency estimation, which is the difference between the vibration main frequency estimation before and after, represents the time length of one step, represents the vibration main frequency estimation value at the current time, represents the vibration main frequency estimation value at the time of the previous step. The sensor comprises an acceleration sensor and a displacement sensor, the acceleration sensor is used to collect acceleration signals of the vibration system, and the displacement sensor is used to collect displacement signals of the vibration system. ； wherein represents the estimation error of the measurement variable, represents the state vector estimate the first component of which represents the sinusoidal component of the vibration displacement at the current time instant, represents the measurement value at the current time instant; Based on the two errors, the signal-to-noise ratio of the signal is calculated The expression is: ； wherein denotes the signal-to-noise ratio of the signal at the current time instant, the numerator in the above equation denotes the accumulated sum of squared estimation errors of the measurement variable, and the denominator denotes the accumulated sum of squared estimation errors of the dominant frequency, reflecting the relative strength between the measurement noise and the system process noise; Let the process noise variance matrix is a known constant matrix, then the measurement noise variance matrix at the current time instant is adaptively adjusted according to the following formula: ； wherein, represents a measurement noise variance matrix at the current time, represents a process noise variance matrix, represents a signal-to-noise ratio of the signal at the current time.

7. A variable impedance vibration control system based on vibration dominant frequency estimation implementing the method according to any one of claims 1 to 6, characterized in that, The system includes a sensor, a processor, a driver, a motor, and a transmission mechanism. The sensor collects vibration signals from the vibration system and sends these signals to the processor. The processor stores a preset frequency band parameter mapping table. It processes the collected vibration signals using signal conditioning algorithms, extended Kalman filtering algorithms, and adaptive noise matrix adjustment algorithms to obtain an estimated dominant vibration frequency. Based on this estimated frequency and the frequency band parameter mapping table, it determines the corresponding impedance parameters, including the impedance stiffness coefficient. Impedance damping coefficient and impedance inertia coefficient Based on the impedance parameters and the displacement difference of the vibration system, ,speed and acceleration Information, through formulas Calculated target control force The processor, based on the target control force Calculate the target current command for the motor. The driver is based on the target current command. The motor is driven to output torque, and the transmission mechanism converts the motor torque into the target control force that matches the vibration direction of the vibration system. It is applied to the vibration system to achieve real-time frequency-domain adaptive vibration control.

8. The variable impedance vibration control system based on vibration dominant frequency estimation of claim 7, wherein, ​ 9. The variable impedance vibration control system based on vibration dominant frequency estimation of claim 7, wherein, The processor comprises a signal conditioning module, a frequency domain estimation module, an impedance mapping module, a force generation module, and a current generation module; the signal conditioning module is configured to receive acceleration sensor signals and displacement sensor signals, perform filtering and smoothing conditioning operations on the signals, and output conditioned sensor signals; the frequency domain estimation module is configured to output real-time vibration main frequency estimation values ; the impedance mapping module is configured to output impedance parameters according to a preset frequency band parameter mapping table; the force generation module is configured to calculate target control forces based on selected impedance parameters and output the target control forces; and the current generation module is configured to receive the target control forces , calculate target current instructions , and output the target current instructions.

10. The variable impedance vibration control system based on vibration dominant frequency estimation of claim 9, wherein, The frequency domain estimation module further comprises an extended Kalman filter module and an adaptive noise matrix adjustment module, the adaptive noise matrix adjustment module is configured to receive an estimated state at a previous time and a measurement signal at a current time, and output an estimated noise matrix at the current time, and the extended Kalman filter module is configured to receive the measurement signal at the current time and the estimated noise matrix, and output a vibration main frequency estimation value at the current time .

Images