A frequency superposition-based online error calibration method for a hemispherical resonator gyro head

Abstract

The application discloses an online error calibration method for a hemispherical resonator gyro based on frequency superposition, which solves the problem that the traditional error calibration method for a hemispherical resonator gyro cannot realize online calibration of gain error, bias angle error and phase error in the same working mode. The application utilizes high-frequency cosine and sine reference signals to demodulate two-channel detection signals and extract high-frequency components in the detection signals. Subsequently, an associated model is established by using the high-frequency cosine and sine given control signals and the demodulated signals to realize online identification of the gyro error, and the identification result is applied to online calibration of the channel gain error, bias angle error and phase error, thereby enhancing the anti-interference ability of the hemispherical resonator gyro under variable temperature conditions. The application is mainly used for simultaneous online calibration of the channel gain error, bias angle error and phase error.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of inertial technology, and is mainly applied to the online calibration process of a head error of an axisymmetric vibration gyroscope. BACKGROUND

[0002] A hemispherical resonator gyroscope measures angular rate by measuring the Coriolis effect of a resonator standing wave rotating around the central axis, and is widely used in land, space, marine, aviation and other fields, and has the advantages of small size, high precision, long service life and low power consumption. A thin chromium gold film is plated between the lower end surface of the hemispherical resonator and the electrode base, which together constitute the detection and excitation capacitance of the hemispherical resonator. Then the electrodes are connected to a buffer amplifier circuit, which converts the capacitance change caused by the vibration of the resonator into a voltage change. With the change of the environmental temperature, the assembly level of the hemispherical resonator gyroscope will change, further causing the gain error and the bias angle error of the X / Y channels of the gyroscope to change. In addition, the performance of the electronic components in the buffer amplifier circuit will also change randomly with the change of the temperature, which causes the transfer function of the detection circuit of the gyroscope to change, and the resonant frequency of the resonator will also change regularly with the temperature, which further causes the phase error of the detection signal at the resonant frequency to change.

[0003] There are separate online calibration methods for the gain error, the bias angle error and the phase error of the X / Y channels in the prior art. The online calibration methods for the gain error and the bias angle error are realized by constructing a virtual precession control loop, and the gain error and the bias angle error of the two channels are identified by the relationship between the standing wave azimuth angle and the amplitude control force and the slow variable signal. However, this method increases the complexity of the control system, and the calibration process causes sinusoidal fluctuations in the main standing wave amplitude, the quadrature wave amplitude and the target phase of the phase-locked loop of the resonator. The online calibration method for the phase error needs to construct the transfer function of the resonator and its supporting control circuit, and uses the amplitude-phase characteristic curve of the driving mode to obtain the phase error of the control loop. This method needs to obtain the amplitude-phase characteristic curve first, and then obtain the phase error according to the resonant frequency, so the calculation process is complex and the environmental adaptability is poor.

[0004] The three errors of gain error, bias angle error and phase error of X / Y channels in the prior art cannot be simultaneously calibrated online, since the prior art identifies and calibrates the gain error and the bias angle error by establishing a detection and drive model of the hemispherical resonator gyro, and reads and compensates the phase error by establishing a transfer function of the hemispherical resonator gyro. The online calibration process of the gain error and the bias angle error requires the gyro to work in a full angle mode and depends on a virtual precession control loop, while the online calibration of the phase error requires the gyro to work in an open loop state. The above calibration schemes have different requirements for the working mode of the gyro, and therefore, the conventional error calibration method of the hemispherical resonator gyro cannot realize online calibration of the gain error, the bias angle error and the phase error in the same working mode. SUMMARY

[0005] The present application aims to solve the problem that the conventional error calibration method of the hemispherical resonator gyro cannot realize online calibration of the gain error, the bias angle error and the phase error in the same working mode, and provides an online calibration method of a hemispherical resonator gyro based on frequency superposition.

[0006] An online calibration method of a hemispherical resonator gyro based on frequency superposition, comprising the following steps:

[0007] Step 1: power on the hemispherical resonator gyro to make the hemispherical resonator gyro in a stable working state;

[0008] Step 2: determine whether the power-on starting time reaches a preset time for starting online calibration of the gyro, and if yes, jump to step 3, otherwise, jump to step 1;

[0009] Step 3: use high-frequency sine reference signal V hs and high-frequency cosine reference signal V hc to demodulate the X channel detection signal x h and the Y channel detection signal y h output by the pre-buffering amplification circuit in the matched control circuit connected with the hemispherical resonator gyro without virtual precession control, thereby obtaining X channel high-frequency component cosine signal C hx , Y channel high-frequency component cosine signal C hy , X channel high-frequency component sine signal S hx and Y channel high-frequency component sine signal S hy .

[0010] wherein, the frequency of V hc and V hs is ω h , ω h >> ω, and ω is the resonant frequency of the resonator.

[0011] Step 4: The k-th round of identification, with k initially set to 1, and the initial value of the meter error set to 0. The amplitude U of the high-frequency cosine given control signal of channel X is set... xc The amplitude U of the high-frequency cosine given control signal for the Y channel yc The amplitude U of the high-frequency sinusoidal given control signal for the X channel xs The amplitude U of the high-frequency sinusoidal given control signal for the Y channel ys As an observable, the C-value demodulated at high frequency hx C hy S hx and S hy As the desired output, a header error identification algorithm is used to identify the header error online; the header error includes the channel gain error k. yx , Angle error α and phase error

[0012] Step 5: Determine whether the header error identified in the kth round has converged. If yes, proceed to step 6; otherwise, k = k + 1 and proceed to step 4.

[0013] Step 6: Utilize the channel gain error k identified in the kth round yx The detection signals of the X and Y channels of the hemispherical resonant gyroscope, as well as the control signals of the X and Y channels of the hemispherical resonant gyroscope, are compensated by the skew angle error α. The phase error identified in the kth round is also used as the target phase-locked phase of the phase-locked loop in the matching control circuit without virtual precession control connected to the hemispherical resonant gyroscope, thus completing the online calibration of the error of the hemispherical resonant gyroscope meter.

[0014] Preferably, the condition for determining whether the header error identified in the kth round has converged in step 5 is:

[0015] The channel gain error k obtained from the identification in round k and round (k-1) yx The difference between the angle error α obtained from the k-th round and the (k-1)-th round, and the phase error obtained from the k-th round and the (k-1)-th round. If the differences between the table headers are all less than the corresponding convergence threshold, the table header error obtained in the kth round of identification is considered to have converged; otherwise, the table header error obtained in the kth round of identification is not converged.

[0016] Preferably, the detection signal x of the X channel is... h and the detection signal y of the Y channel h They are respectively:

[0017]

[0018] X-channel high-frequency component cosine signal C hx Y-channel high-frequency component cosine signal C hyX channel high frequency component sine signal S hx and Y channel high frequency component sine signal S hy The expression is:

[0019]

[0020]

[0021]

[0022]

[0023] Wherein, LPF(·) is a low-pass filtering operation, k x is the X channel detection gain, k y is the Y channel detection gain, x hc and y hc are high frequency cosine components of the X channel and Y channel detection signals respectively, x hs and y hs are high frequency sine components of the X channel and Y channel detection signals respectively, a represents the amplitude of the main standing wave, q represents the amplitude of the quadrature wave, k x is the X channel detection gain, k y is the Y channel detection gain, t is time, θ is the azimuth angle of the standing wave, is the X channel detection signal x h and Y channel detection signal y h output by the pre-buffering amplification circuit, and the phase difference between the detection signals before amplification of the corresponding channels at the frequency ω h .

[0024] Preferably, the implementation of the online identification of the table head error in step 4 by using the table head error identification algorithm includes:

[0025] Step 41, according to U xc , U yc , U xs and U ys as the observation quantities, and C hx , C hy , S hx and S hy as the expected output quantities, a correlation model is established:

[0026]

[0027] Wherein,

[0028]

[0029]

[0030] In the above formula, A is the detection error matrix, E is the driving error matrix, D is the phase error matrix, and k x For the detection gain of the X channel, k y For the Y-channel detection gain, K x K controls the gain of channel X. y For Y-channel control gain, The detection signal x of the X channel output by the pre-buffer amplifier circuit h and the detection signal y of the Y channel h The detection signal before amplification of its corresponding channel at frequency ω h The phase difference at that point, and

[0031] Step 42: Convert the correlation model into a discrete product form:

[0032] C(k) = U(k)β(k);

[0033] in:

[0034] C(k)=[S hx (k) S hy (k) C hx (k) C hy (k)] T ;

[0035]

[0036]

[0037] C(k) is the expected output vector at the k-th sampling time, U(k) is the observation vector at the k-th sampling time, β(k) is the parameter vector at the k-th sampling time, β1(k), β2(k), β3(k), and β4(k) are the elements in the first column of the first to fourth rows of β(k), respectively, and k is the sampling time sequence number, k = 1, 2, 3, ..., C hx (k) is C hx The discrete form of C hy (k) is C hy The discrete form of S hx (k) is S hx The discrete form of S hy (k) is S hy The discrete form of U xc (k) is U xc The discrete form of U xs (k) is U xs The discrete form of U yc (k) is U yc The discrete form of U ys (k) is Uys a discrete form of k x (k) is a discrete form of k x a discrete form of K x (k) is a discrete form of K x a discrete form of k y (k) is a discrete form of k y a discrete form of K y (k) is a discrete form of K y a discrete form of a is a discrete form of ;

[0038] Step 43, according to U(k) in step 42, the gain vector H(k) and the covariance matrix P(k) of the table head error identification algorithm corresponding to the kth sampling time are obtained;

[0039] Step 44, according to C(k) and U(k) obtained in step 42, and H(k) obtained in step 43, the parameter vector β(k) of the kth sampling time is obtained;

[0040] Step 45, according to ω h , ω, and β1(k), β2(k), β3(k) and β4(k) in β(k), the channel gain error k yx , the angle error a and the phase error of the kth round of identification are obtained, so as to complete the online identification of the table head error;

[0041]

[0042]

[0043]

[0044] wherein,

[0045] Preferably, in step 43,

[0046]

[0047]

[0048] wherein, P(0) = I, I is a unit matrix, η is a forgetting factor, η is selected in the range of (0, 1], and P(k-1) is the covariance matrix corresponding to the k-1th sampling time.

[0049] Preferably, in step 44, β(k) = β(k-1) + H(k)(C(k)-U(k)β(k-1));

[0050] Wherein, β(k-1) is the parameter vector of the k-1th sampling time,

[0051] Preferably, the channel gain error k yx and the bias angle error α are used to compensate the detection signals of the X channel and the Y channel of the hemispherical resonator gyroscope in step 6.

[0052]

[0053] x com is the compensated detection signal of the X channel of the hemispherical resonator gyroscope, and y com is the compensated detection signal of the Y channel of the hemispherical resonator gyroscope.

[0054] Preferably, the channel gain error k yx and the bias angle error α are used to compensate the control signals of the X channel and the Y channel of the hemispherical resonator gyroscope in step 6.

[0055] Firstly, V hs , V hc , U xc , U yc , U xs and U ys are superimposed on the X channel control signal V x and the Y channel control signal V y to obtain the high-frequency control signal V x1 of the X channel and the high-frequency control signal V y1 of the Y channel; wherein,

[0056] Secondly, the channel gain error k yx and the bias angle error α are used to compensate V x1 and V y1 to obtain the compensated control signal V xcom of the X channel and the compensated control signal V ycom of the Y channel.

[0057]

[0058] Preferably, the stable working state in step 1 is that the amplitude of the resonator of the hemispherical resonator gyroscope is kept constant, and the orthogonal wave of the resonator is zero.

[0059] Preferably, ω h > 10ω.

[0060] Advantages of the present application:

[0061] The online error calibration method of the half-sphere resonator gyro head based on frequency superposition can superimpose a high-frequency reference signal in a control signal, and extract gain error, bias angle error and phase error from a high-frequency component of a resonator X or Y channel detection signal, thereby solving the problem that the prior art cannot simultaneously calibrate the three errors online.

[0062] The method is not limited to the working mode of the gyro or the specific structure of the resonator, and thus has wide applicability, and is particularly suitable for the online error calibration process of the axisymmetric vibration gyro head.

[0063] The method not only improves the detection accuracy of the gyro, but also provides an efficient and flexible new method for error calibration of the axisymmetric vibration gyro, and has important significance for improving the overall performance and application range of the gyro technology. BRIEF DESCRIPTION OF DRAWINGS

[0064] Figure 1 is a structural schematic diagram of a control circuit for driving the half-sphere resonator gyro without virtual precession in the prior art; wherein x h and y h are X channel and Y channel detection signals of the half-sphere resonator gyro, that is, signals amplified by a pre-buffering amplification circuit from vibration signals output by the X channel and Y channel of the half-sphere resonator gyro; V s and V c are sine and cosine reference signals output by a phase-locked loop, E is the total energy of the gyro, so as to keep the vibration amplitude of the resonator constant; Q is the quadrature of the gyro, and is used to eliminate the phase difference between the two channels; L is a phase-locked sine control quantity, and M is a phase-locked cosine control quantity, and the phase difference between the reference signal and the detection signal generated by the vibration of the resonator can be obtained by taking the inverse tangent of the ratio of L and M, and the phase difference is taken as the control quantity of the phase-locked loop; C x , C y , S x , S y are four slowly varying signals, V x and V y are control signals of the X channel and the Y channel.

[0065] Figure 2 is a principle schematic diagram after the online error calibration method of the half-sphere resonator gyro head based on frequency superposition is introduced in Figure 1

[0066] Figure 3 ​is a flow chart of an error online calibration method of a hemispherical resonator gyro based on frequency superposition. DETAILED DESCRIPTION

[0067] The technical solutions in the embodiments of the present application will be clearly and completely described with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0068] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0069] DETAILED DESCRIPTION Figure 2 and Figure 3 The present embodiment describes an error online calibration method of a hemispherical resonator gyro based on frequency superposition. The method comprises the following steps:

[0070] Step 1: power on the hemispherical resonator gyro to make the hemispherical resonator gyro in a stable working state. In a specific application, the stable working state is that the amplitude of the resonator of the hemispherical resonator gyro remains constant, and the orthogonal wave of the resonator is zero.

[0071] Step 2: determine whether the power-on start time reaches the preset time for starting the online calibration of the error of the watch head. If the result is yes, go to step 3, otherwise, go to step 1. The preset time for starting the online calibration of the error of the watch head is set as a time threshold, which is to further ensure that the hemispherical resonator gyro is in a stable working state, and to provide accurate data basis for subsequent accurate calibration.

[0072] Step 3: demodulate the two-channel detection signals by using high-frequency sine and cosine reference signals to realize the extraction of high-frequency components in the detection signals: use high-frequency sine reference signal V hs and high-frequency cosine reference signal V hc to demodulate the X-channel detection signal x h and Y-channel detection signal y h output by the pre-buffering amplification circuit in the matched control circuit connected with the hemispherical resonator gyro without virtual precession control, so as to obtain X-channel high-frequency component cosine signal C hx , Y-channel high-frequency component cosine signal C hy , X-channel high-frequency component sine signal S hx and Y-channel high-frequency component sine signal S hy .

[0073] Wherein, V hc and Vhs The frequency of each of the above is ω h , ω h > ω, ω is the resonance frequency of the resonator; in specific applications, ω h > 10ω to avoid the influence of high-frequency signals on the resonance amplitude;

[0074] Step 4, the initial value of k is 1, the initial value of the table head error is set to 0, the amplitude U xc of the X-channel high-frequency cosine given control signal, the amplitude U yc of the Y-channel high-frequency cosine given control signal, the amplitude U xs of the X-channel high-frequency sine given control signal, and the amplitude U ys of the Y-channel high-frequency sine given control signal are used as observation quantities, C hx , C hy , S hx , and S hy demodulated from the high frequency are used as expected output quantities, and a table head error on-line identification algorithm is used to realize on-line identification of the table head error; the table head error includes channel gain error k yx , bias angle error α, and phase error

[0075] Step 5, whether the table head error identified in the kth round of identification converges is judged, and if yes, the process jumps to step 6, otherwise, k=k+1, and the process jumps to step 4;

[0076] Step 6, the channel gain error k yx and the bias angle error α identified in the kth round of identification are used to compensate the detection signals of the X channel and the Y channel of the hemispherical resonator gyroscope and the control signals of the X channel and the Y channel of the hemispherical resonator gyroscope, and the phase error identified in the kth round of identification is used as the target phase-locked phase of the phase-locked loop in the matched control circuit connected with the hemispherical resonator gyroscope without virtual precession control, to complete on-line calibration of the table head error of the hemispherical resonator gyroscope.

[0077] In specific applications, the control signals of the X channel and the Y channel of the hemispherical resonator gyroscope are used to drive the excitation electrodes of the X channel and the Y channel of the hemispherical resonator gyroscope, and under the action of the control signals of the X channel and the Y channel, the detection electrodes of the X channel and the Y channel of the hemispherical resonator gyroscope output detection signals, which are amplified by the pre-buffering amplification circuit in the matched control circuit connected with the hemispherical resonator gyroscope without virtual precession control, as shown in Figure 1 .

[0078] The frequency superposition in the on-line calibration method of the table head error of the hemispherical resonator gyroscope based on frequency superposition according to the embodiment refers to the superposition of the resonance frequency and the high frequency in the two-channel detection signals and control signals.

[0079] Furthermore, the condition for determining whether the header error identified in the kth round has converged in step 5 is:

[0080] The channel gain error k obtained from the identification in round k and round (k-1) yx The difference between the angle error α obtained from the k-th round and the (k-1)-th round, and the phase error obtained from the k-th round and the (k-1)-th round. If the differences between the table headers are all less than the corresponding convergence threshold, the table header error obtained in the kth round of identification is considered to have converged; otherwise, the table header error obtained in the kth round of identification is not converged.

[0081] See Figure 2 Furthermore, the detection signal x of the X channel h and the detection signal y of the Y channel h They are respectively:

[0082]

[0083] X-channel high-frequency component cosine signal C hx Y-channel high-frequency component cosine signal C hy X-channel high-frequency component sinusoidal signal S hx and the high-frequency component sinusoidal signal S of the Y channel hy The expression is:

[0084]

[0085]

[0086]

[0087]

[0088] Where LPF(·) is the low-pass filter operation, k x For the detection gain of the X channel, k y For the Y-channel detection gain, x hc and y hc These are the high-frequency cosine components of the detection signals from the X and Y channels, respectively. hs and y hs These are the high-frequency sinusoidal components of the detection signals from the X and Y channels, respectively. 'a' represents the amplitude of the main standing wave, 'q' represents the amplitude of the orthogonal wave, and 'k' represents the amplitude of the orthogonal wave. x For the detection gain of the X channel, k y The Y-channel detection gain is given by t, where t is time and θ is the azimuth angle of the standing wave. The detection signal x of the X channel output by the pre-buffer amplifier circuit h and the detection signal y of the Y channel h The detection signal before amplification of its corresponding channel at frequency ωh the phase difference between the detection signal and the corresponding channel detection signal before amplification at the frequency ω

[0089] Since the hemispherical resonator gyroscope is a linear system, superimposing a high-frequency signal on the control signal will cause the resonator to produce a high-frequency response, and the size of the high-frequency response is independent of the position of the standing wave azimuth angle, so this method is not affected by the working model of the gyroscope and the structure shape of the resonator. At the same time, the changes of the gain error, the bias angle error and the phase error of the hemispherical resonator gyroscope will cause the response of the resonator to the high-frequency control signal to change, so a correlation model can be established between the amplitude of the given high-frequency signal and the size of the high-frequency component in the resonator detection signal, which covers the gain error, the bias angle error and the phase error of the gyroscope and can realize the subsequent simultaneous identification of the header error. Therefore, the online identification of the header error can be realized by constructing the correlation model.

[0090] Further, the implementation manner of step 4 for realizing the online identification of the header error by using the header error identification algorithm comprises:

[0091] Step 41, a correlation model is established according to U xc , U yc , U xs and U ys as observation quantities and C hx , C hy , S hx and S hy as expected output quantities.

[0092]

[0093] wherein,

[0094]

[0095]

[0096] In the above formula, A is a detection error matrix, E is a driving error matrix, D is a phase error matrix, k x is the X-channel detection gain, k y is the Y-channel detection gain, K x is the X-channel control gain, K y is the Y-channel control gain, is the phase difference between the detection signal x h and the detection signal y h of the X channel and the Y channel output by the pre-buffering amplification circuit and the corresponding channel detection signal before amplification at the frequency ω h , and

[0097] Step 42, converting the correlation model into discrete product form:

[0098] C(k) = U(k) β(k);

[0099] wherein:

[0100] C(k) = [S hx (k) S hy (k) C hx (k) C hy (k)] T ;

[0101]

[0102]

[0103] T is matrix transpose, C(k) is the expected output vector at the kth sampling time, U(k) is the observation vector at the kth sampling time, β(k) is the parameter vector at the kth sampling time, β1(k), β2(k), β3(k) and β4(k) are the first column elements of the first to fourth rows in β(k) respectively, k is the sampling time sequence number, k = 1, 2, 3, …, C hx (k) is the discrete form of C hx , C hy (k) is the discrete form of C hy , S hx (k) is the discrete form of S hx , S hy (k) is the discrete form of S hy , U xc (k) is the discrete form of U xc , U xs (k) is the discrete form of U xs , U yc (k) is the discrete form of U yc , U ys (k) is the discrete form of U ys , k x (k) is the discrete form of k x , K x (k) is the discrete form of K x , k y (k) is the discrete form of k y , K y (k) is the discrete form of K y , α(k) is the discrete form of α, is the discrete form of ;

[0104] Step 43, according to U(k) in step 42, the gain vector H(k) and the covariance matrix P(k) of the table head error identification algorithm corresponding to the k-th sampling time are obtained;

[0105]

[0106]

[0107] wherein P(0) = I, I is a unit matrix, η is a forgetting factor, η is selected in the range of (0, 1], P(k-1) is the covariance matrix corresponding to the k-1-th sampling time;

[0108] Step 44, according to C(k) and U(k) obtained in step 42, and H(k) obtained in step 43, the parameter vector β(k) of the k-th sampling time is obtained;

[0109] β(k) = β(k-1) + H(k)(C(k) - U(k)β(k-1));

[0110] wherein β(k-1) is the parameter vector of the k-1-th sampling time, H(k) reflects the influence of the error on β(k);

[0111] Step 45, according to ω h , ω, and β1(k), β2(k), β3(k) and β4(k) in β(k), the channel gain error k yx , the angle error α and the phase error of the k-th round of identification are obtained;

[0112]

[0113]

[0114]

[0115] wherein,

[0116] Further, the implementation mode of compensating the detection signals of the X channel and the Y channel of the hemispherical resonator gyro by using the channel gain error k yx and the angle error α identified in the k-th round of identification in step 6 is as follows:

[0117]

[0118] x com is the compensated detection signal of the X channel of the hemispherical resonator gyro, y com is the compensated detection signal of the Y channel of the hemispherical resonator gyro.

[0119] Further, the channel gain error k yx and the bias angle error α are used to compensate the control signals of the X channel and the Y channel of the hemispherical resonator gyroscope in step 6.

[0120] Firstly, V hs , V hc , U xc , U yc , U xs and U ys are superimposed on the control signal V x of the X channel and the control signal V y of the Y channel to obtain the high-frequency control signal V x1 of the X channel and the high-frequency control signal V y1 of the Y channel; wherein,

[0121]

[0122] Secondly, the channel gain error k yx and the bias angle error α are used to compensate V x1 and V y1 to obtain the compensated control signal V xcom of the X channel and the compensated control signal V ycom of the Y channel.

[0123]

[0124] Principle analysis:

[0125] The application provides an online calibration method for a hemispherical resonator gyroscope based on frequency superposition, which superimposes high-frequency sine and cosine signals on the control signals of amplitude and quadrature control, demodulates two-channel detection signals by using high-frequency sine and cosine reference signals, and extracts high-frequency components in the detection signals. Subsequently, the online identification of the head error is realized by using the correlation model established by the high-frequency sine and cosine given control signals of the two channels and the demodulated signals, and the identification results are applied to the online calibration of the channel gain error, the bias angle error and the phase error, so that the anti-interference ability of the hemispherical resonator gyroscope under variable temperature conditions is enhanced. The calibration scheme does not need additional experimental equipment, but only relies on the hemispherical resonator gyroscope and the matching control circuit without virtual precession control, see Figure 1 and Figure 2 Therefore, the application can be widely applied to the field of online calibration of the hemispherical resonator gyroscope head error.

[0126] While the application has been described with reference to particular embodiments thereof, it is to be understood that these embodiments are merely illustrative of the principles and applications of the present application. It will be apparent to those skilled in the art that numerous modifications can be made within the scope of the present application as defined by the appended claims. It is intended that all such modification fall within the spirit and scope of the present application. It will be understood that the features described in connection with one embodiment can be used in connection with another embodiment.

Claims

1. A frequency superposition-based on-line error calibration method for a hemispherical resonator gyro head, characterized in that, The method comprises the following steps: Step 1, power-on start the hemispherical resonator gyroscope, so that the hemispherical resonator gyroscope is in a stable working state; Step 2, judge whether the power-on start time reaches the preset time of opening the watch head error online calibration, the result is yes, jump to step 3, otherwise, jump to step 1; Step 3, high frequency sine reference signal and high frequency cosine reference signal The detection signal of X channel outputted by the pre-buffering amplification circuit in the matched control circuit connected with the hemispherical resonator gyro without virtual precession control and the detection signal of Y channel High frequency demodulation is carried out, so as to obtain X channel high frequency component cosine signal , Y channel high frequency component cosine signal , X channel high frequency component sine signal and Y channel high frequency component sine signal ; wherein and the frequency of each of , , is the resonance frequency of the resonator. Step 4, the initial value of the step 3 wheel recognition, The initial value of the table head error is 0, and the amplitude of the X channel high-frequency cosine given control signal is set as 1 , the amplitude of the Y channel high-frequency cosine given control signal , the amplitude of the X channel high-frequency sine given control signal and the amplitude of the Y channel high-frequency sine given control signal As an observation quantity, the high-frequency demodulated , , and As an expected output quantity, the table head error is realized by using a table head error identification algorithm for online identification of the table head error; the table head error includes channel gain error , bias angle error and phase error ; The implementation mode of realizing the online identification of the watch head error by using the watch head error identification algorithm comprises: Step 41, establish a correlation model based on , , and as observation quantities, and , , and as desired output quantities. ； Wherein, In the above formula, is a detection error matrix, is a drive error matrix, is a phase error matrix, is an X channel detection gain, is a Y channel detection gain, is an X channel control gain, is a Y channel control gain, is a phase difference between the detection signal of the X channel output by the pre-buffering amplification circuit and the detection signal of the Y channel and the detection signal of the corresponding channel before amplification at a frequency , and ; Step 42, convert the correlation model into a discrete product form: ； Wherein: For the first The expected output vector at each sampling time point, For the first The observation vector at each sampling time, For the first The parameter vector at each sampling time point , , and They are respectively The first column of the first to fourth rows, This is the sampling time sequence number. , for discrete form, for discrete form, for discrete form, for discrete form, for discrete form, for discrete form, for discrete form, for discrete form, for discrete form, for discrete form, for discrete form, for discrete form, for discrete form, for Discrete form; Step 43, according to step 42 , the gain vector and the covariance matrix of the table head error identification algorithm corresponding to the first sampling moment are obtained ; Step 44, the parameter vector at the first sampling time is calculated according to the parameter vector obtained in step 42 and the parameter vector obtained in step 43 . ​​ Step 45, according to , , and in , , and , get the identified channel gain error , the angle error and the phase error of the first round, complete the online identification of the table head error; wherein ; Step 5, determine the first If the identified header error has converged, proceed to step 6; otherwise, check if the error has converged. Proceed to step 4; Step 6, using the first Wheel-identified channel gain error And the error of the deflection angle The detection signal of the X channel and the Y channel of the hemispherical resonator gyroscope, and the control signal of the X channel and the Y channel of the hemispherical resonator gyroscope are compensated, and the phase error identified by the first Wheel is used as the target phase-locked phase of the phase-locked loop in the matched control circuit connected with the hemispherical resonator gyroscope without virtual precession control, and the online calibration of the head error of the hemispherical resonator gyroscope is completed.

2. The frequency superposition based on-line error calibration method for a hemispherical resonator gyro head according to claim 1, characterized in that, The condition for determining whether the table head error recognized by the wheel is converged in step 5 is: whether the table head error recognized by the wheel is converged in step 5 is: First Wheel and the first Channel gain error obtained by wheel recognition Interval difference value, the first Wheel and the first Angle error obtained by wheel recognition Interval difference value, and the first Wheel and the first Phase error obtained by wheel recognition When the interval difference value is less than the corresponding convergence threshold value, it is determined that the table head error obtained by the first Wheel recognition converges, otherwise, the table head error obtained by the first Wheel recognition does not converge.

3. The frequency superposition based on-line error calibration method for a hemispherical resonator gyro according to claim 1, characterized in that, the detection signal of the x channel the detection signal of the y channel respectively ； X channel high frequency component cosine signal , Y channel high frequency component cosine signal , X channel high frequency component sine signal , and Y channel high frequency component sine signal The expression is: ； in, This is a low-pass filter operation. For X channel detection gain, For Y-channel detection gain, and These are the high-frequency cosine components of the detection signals from the X and Y channels, respectively. and These are the high-frequency sinusoidal components of the detection signals from the X and Y channels, respectively. Represents the amplitude of the main standing wave. Represents the amplitude of orthogonal waves. For X channel detection gain, The Y-channel detection gain is given by t, where t is time and θ is the azimuth angle of the standing wave. The detection signal of the X channel output by the pre-buffer amplifier circuit and the detection signal of the Y channel The detection signal before amplification of its corresponding channel at the frequency The phase difference at that point.

4. The error online calibration method of a frequency superposition based hemispherical resonator gyroscope based on the method of claim 1, characterized in that, In step 43, ； ； wherein , is an identity matrix, is a forgetting factor, is selected in the range (0, 1], is the covariance matrix corresponding to the th sampling instant.

5. The error online calibration method of a frequency superposition based hemispherical resonator gyroscope based on the method of claim 1, characterized in that, In step 44, ; wherein is the parameter vector at the th sampling instant, .

6. The frequency superposition based on-line error calibration method for a hemispherical resonator gyro according to claim 1, wherein The first Wheel-identified channel gain error And the angle error The implementation of compensating the detection signals of the X channel and the Y channel of the hemispherical resonator gyro is as follows: ； the detection signal of the X channel of the compensated hemispherical resonator gyroscope, the detection signal of the Y channel of the compensated hemispherical resonator gyroscope.

7. The error online calibration method of a frequency superimposition based hemispherical resonator gyroscope based on the method of claim 1, wherein, The first Wheel-identified channel gain error And the angle error The implementation mode of compensating the control signals of the X channel and the Y channel of the hemispherical resonator gyro comprises: Firstly, superimpose , , , , and to the control signals of the X channel and the control signals of the Y channel to obtain the high-frequency control signals of the X channel and the high-frequency control signals of the Y channel ; wherein, ； Second, reuse the first Wheel recognized channel gain error And the angle error To And Compensation, get compensated X channel control signal And the compensation of Y channel control signal ; 。 8. The frequency superposition based on-line error calibration method for hemispherical resonator gyro (HRG) according to claim 1, wherein, The stable working state in step 1 is that the resonator amplitude of the hemispherical resonator gyroscope remains constant, and the resonator orthogonal wave is zero.

9. The error online calibration method of a frequency superimposition based hemispherical resonator gyroscope based on the method of claim 1, wherein, 。

Images