Deep-sea robot electric drive system predistortion fourth-order generalized integral resonance identification method

Abstract

The invention discloses a deep-sea robot electric drive system predistortion fourth-order generalized integral resonance identification method, which comprises the following steps: S1, sampling the rotating speed of a motor, and calculating a rotating speed error signal u between a rotating speed given value and a sampling value; s2, obtaining a signal # imgabs0 # orthogonal to the rotating speed error signal u, and multiplying the signal # imgabs0 # by the rotating speed error signal u to obtain an error signal # imgabs1 #; s3, configuring resonance identification parameters based on high identification precision and strong harmonic suppression performance; s4, discretizing the fourth-order generalized integral element by adopting a pre-distortion discretization method; and step S5, extracting a direct current component # imgabs3 # containing resonance information in the error signal # imgabs2 #, and obtaining a vibration frequency estimated value # imgabs4 # of the rotating speed error signal u. According to the method, online resonance identification with strong anti-interference and high precision can be realized, the resonance observer can quickly configure parameters, and the configured parameters have high identification precision and strong harmonic suppression performance.

Description

Technical Field

[0001] The present invention relates to the field of deep-sea motor control technology, and more specifically, to a method for identifying fourth-order generalized integral resonance pre-distortion of a deep-sea robot electric drive system. Background Art

[0002] Deep-sea robots have perfect sensing and propulsion systems, and complete seabed operations by controlling deep-sea manipulators. For deep-sea manipulators, their drive systems are affected by the agitation of seawater and excite resonance. The resonance factor will cause severe vibration of the deep-sea manipulator, further affecting deep-sea operations and greatly increasing the risk of damage to the deep-sea manipulator. For applications such as deep-sea robots that require high reliability, a high-precision, strong anti-interference online resonance identification method is essential.

[0003] Common online resonance identification methods include strategy groups based on Fourier transform, such as discrete Fourier analysis, fast Fourier analysis, and sliding window discrete Fourier analysis. Compared with the online resonance identification algorithm based on Fourier transform, the online resonance identification algorithm based on the orthogonal principle has the characteristics of fast convergence speed and small amount of calculation. However, the traditional resonance observer based on the orthogonal principle has poor anti-interference performance for harmonics, resulting in low accuracy under harmonic interference. Even if a second-order generalized integrator is used, the anti-interference performance for high-order harmonics can only be improved to a limited extent, but it has no anti-interference effect on low-order harmonics. In addition, in the actual application of the resonance observer, the offset at the center frequency caused by the unmatched discretization method further reduces the identification accuracy of the online resonance observer, resulting in the application of the traditional resonance identification method to deep-sea robots facing the dual problems of poor anti-interference ability and low identification accuracy. Summary of the invention

[0004] The purpose of the present invention is to overcome the deficiencies of the above-mentioned prior art and to provide a pre-distortion fourth-order generalized integral resonance identification method for a deep-sea robot electric drive system so as to suppress the resonance identification harmonics, improve the resonance identification accuracy, and enhance the operation reliability of the deep-sea robot.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] The fourth-order generalized integral resonance identification method for the pre-distortion of the electric drive system of a deep-sea robot includes the following steps:

[0007] Step S1, sampling the motor speed and calculating the speed error signal u between the speed given value and the sampled value;

[0008] Step S2, using a fourth-order generalized integral link to obtain a signal orthogonal to the speed error signal u , and multiply it with the speed error signal u to obtain the error signal ;

[0009] Step S3, configuring resonance identification parameters based on high identification accuracy and strong harmonic suppression performance;

[0010] Step S4, using the pre-distortion discretization method to discretize the fourth-order generalized integral link to improve the error signal The accuracy of

[0011] Step S5, extracting the error signal using a low-pass filter The DC component containing the resonance information The vibration frequency estimate of the speed error signal u is obtained through the integral controller , estimated vibration frequency Used for resonance suppression of electric drive systems of deep-sea robots.

[0012] Furthermore, in step S2, the error signal Make judgments on low-order and high-order harmonic interference to facilitate subsequent resonance identification parameter configuration.

[0013] Furthermore, the low-order harmonic interference determination includes the following steps:

[0014] Step S201, design a fourth-order generalized integral link:

[0015] (1)

[0016] In formula (1), is the transfer function of the fourth-order generalized integral link, j is the imaginary unit, is the angular frequency; and is the coefficient; is the estimated value of vibration frequency; is the amplitude of the fourth-order generalized integral link; is the phase angle of the fourth-order generalized integral link;

[0017] Step S202: Based on formula (1), the amplitude of the low-order harmonics after conditioning by the fourth-order generalized integral link is ,because is an estimate of the vibration frequency, so Close to zero, which means the error signal No low-order harmonic interference;

[0018] Furthermore, the high-order harmonic interference judgment includes the following steps:

[0019] Step S203, using a fourth-order generalized integral link to obtain a signal orthogonal to the speed error signal u It can be expressed as:

[0020] (2)

[0021] In formula (2), is the fundamental frequency amplitude of the fourth-order generalized integral link; is the fundamental frequency phase angle of the fourth-order generalized integral link; is the fundamental angular frequency of the fourth-order generalized integral link; is the k-order harmonic amplitude of the fourth-order generalized integral link; is the kth order harmonic phase angle of the fourth order generalized integral link; is the kth order harmonic angular frequency of the fourth order generalized integral link;

[0022] Step S204, based on formula (2), after the high-order harmonics are conditioned by the fourth-order generalized integral link, their amplitude decreases as the harmonic order increases, and their influence is negligible, which means that the error signal No high-order harmonic interference;

[0023] Furthermore, the error signal It is expressed as:

[0024] (3)

[0025] In formula (3), is the amplitude of the DC component in the speed error signal u; is the amplitude of the fundamental frequency component in the speed error signal u; is the amplitude of the higher harmonic components in the speed error signal u.

[0026] Furthermore, in step S3, the configuration of the resonance identification parameters includes the following steps:

[0027] Step S301, high recognition accuracy performance condition: transfer function The phase-frequency characteristic at the center frequency shows a phase lead of 90 degrees to ensure the convergence accuracy of the fourth-order generalized integral link to the orthogonal frequency;

[0028] Strong harmonic suppression performance conditions: transfer function The amplitude gain at the center frequency is 1, and the amplitude at frequencies other than the center frequency is significantly attenuated to ensure that the fourth-order generalized integral link suppresses low-order and high-order harmonics;

[0029] Step S302: Based on the conditions of high identification accuracy and strong harmonic suppression performance, set the coefficient The initial value is 1, and the coefficient is given The transfer function changes from 0.1 to 1 in 0.1 steps. Bode diagram;

[0030] Step S303, fixed coefficient The value of remains unchanged, based on the rapidity and stability of the center frequency convergence, the coefficient The value of is adjusted to 0.3;

[0031] Step S304, fixed coefficient The value of is 0.3, giving the coefficient The transfer function varies from 0.2 to 2 Bode plot, based on the harmonic suppression capability, the coefficients The value of is adjusted to 1.4.

[0032] Furthermore, step S4 specifically includes the following steps:

[0033] In step S401, the pre-distortion discretization method used is expressed as:

[0034] (4)

[0035] In formula (4), To control the cycle;

[0036] Step S402: discretized fourth-order generalized integral link and its phase angle It is expressed as:

[0037]

[0038] = (4)

[0039] (5)

[0040] = (6)

[0041] In formula (6), for .

[0042] Furthermore, step S5 specifically includes the following steps:

[0043] Step S501: extract the error signal using a low-pass filter The DC component containing the resonance information , expressed as:

[0044] (7)

[0045] Step S502: Using an integral regulator to adjust the DC component Set it to zero, and we get:

[0046] (8)

[0047] Will Converge to , we can get the estimated value of the vibration frequency of u .

[0048] The beneficial effects of the present invention are:

[0049] 1. The present invention can realize strong anti-interference and high-precision online resonance identification. The resonance observer can quickly configure parameters, and the configured parameters have high identification accuracy and strong harmonic suppression performance.

[0050] 2. The fourth-order generalized integral link in the present invention can effectively suppress low-order and high-order resonances, and when DC disturbances and high-order harmonic disturbances occur, it can quickly suppress the disturbances and converge back to the accurate value.

[0051] 3. The pre-distortion discretization in the present invention can make the resonance identification method robust to the control frequency and can accurately identify the resonance at a control frequency of 10k to 1k.

[0052] 4. On the premise of having strong anti-interference and high-precision online resonance identification performance, the pre-distortion fourth-order generalized integral resonance identification method of the present invention has the advantages of small computational load and short total computational time. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic diagram of a control structure of a method for identifying a fourth-order generalized integral resonance pre-distortion of an electric drive system of a deep-sea robot in this embodiment;

[0054] Figure 2 The coefficient in this embodiment is A schematic diagram of parameter configuration;

[0055] Figure 3 The coefficient in this embodiment is A schematic diagram of parameter configuration;

[0056] FIG4( a ) is a waveform diagram of the traditional orthogonal principle method under different harmonic disturbances;

[0057] FIG4( b ) is a waveform diagram of the fourth-order generalized integral resonance identification method for pre-distortion of the electric drive system of the deep-sea robot under different harmonic disturbances in this embodiment;

[0058] Figure 5 A waveform comparison diagram of the traditional method and the fourth-order generalized integral resonance identification method for the pre-distortion of the electric drive system of the deep-sea robot of this embodiment at different control frequencies;

[0059] Figure 6 This is a running time comparison chart of the fourth-order generalized integral resonance identification method for pre-distortion of the deep-sea robot electric drive system of the present embodiment in TMS320F28335 according to the conventional method 1 and the conventional method 2. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] Embodiment: A method for identifying a fourth-order generalized integral resonance with pre-distortion in a deep-sea robot electric drive system is used to suppress resonance identification harmonics, improve resonance identification accuracy, and enhance the operating reliability of the deep-sea robot. Figure 1 As shown, the following steps are included:

[0062] Step S1, sampling the motor speed and calculating the speed error signal u between the speed given value and the sampled value;

[0063] In this embodiment, an encoder is used to sample the motor position, and the motor speed is calculated by TMS320F28335 to obtain a speed error signal u between the speed setting value and the sampling value. The speed error signal u is directly used as an effective signal for identifying resonance.

[0064] Step S2, using a fourth-order generalized integral link to obtain a signal orthogonal to the speed error signal u , and multiply it with the speed error signal u to obtain the error signal ;

[0065] Further, in step S2, by obtaining the error signal The low-order and high-order harmonic interference judgment is carried out to facilitate the subsequent resonance identification parameter configuration, including the following steps:

[0066] (1)

[0067] In formula (1), is the transfer function of the fourth-order generalized integral link, j is the imaginary unit, is the angular frequency; and is the coefficient; is the estimated value of vibration frequency; is the amplitude of the fourth-order generalized integral link; is the phase angle of the fourth-order generalized integral link;

[0068] Step S202: Based on formula (1), for low-order harmonics, especially harmonics with frequencies much smaller than the fundamental frequency, the amplitude is conditioned by the fourth-order generalized integral link. ,because is an estimate of the vibration frequency, so Close to zero, which means the error signal No low-order harmonic interference;

[0069] Step S203, using a fourth-order generalized integral link to obtain a signal orthogonal to the speed error signal u It can be expressed as:

[0070] (2)

[0071] In formula (2), is the fundamental frequency amplitude of the fourth-order generalized integral link; is the fundamental frequency phase angle of the fourth-order generalized integral link; is the fundamental angular frequency of the fourth-order generalized integral link; is the k-order harmonic amplitude of the fourth-order generalized integral link; is the kth order harmonic phase angle of the fourth order generalized integral link; is the kth order harmonic angular frequency of the fourth order generalized integral link;

[0072] Step S204, based on formula (2), after the high-order harmonics are conditioned by the fourth-order generalized integral link, their amplitude decreases as the harmonic order increases, and their influence can be ignored, which means that the error signal No high-order harmonic interference;

[0073] Step S205, based on step S202 and step S204, the error signal It is expressed as:

[0074] (3)

[0075] In formula (3), is the amplitude of the DC component in the speed error signal u; is the amplitude of the fundamental frequency component in the speed error signal u; is the amplitude of the high-order harmonic component in the speed error signal u. Can be used for subsequent resonance identification.

[0076] Step S3, configuring resonance identification parameters based on the two conditions of high identification accuracy and strong harmonic suppression performance;

[0077] Furthermore, in step S3, the configuration of the resonance identification parameters includes the following steps:

[0078] Step S301, high identification accuracy performance condition: Since the fourth-order generalized integral link converges to the orthogonal frequency, in order to ensure the convergence accuracy, the transfer function The phase-frequency characteristic at the center frequency should show a phase lead of 90 degrees;

[0079] Strong harmonic suppression performance conditions: In order to ensure that the fourth-order generalized integral link has the ability to suppress low-order and high-order harmonics, the transfer function The amplitude gain at the center frequency should be 1, and there should be significant amplitude attenuation at frequencies other than the center frequency;

[0080] Step S302: Based on the conditions of high identification accuracy and strong harmonic suppression performance, set the coefficient The initial value is 1, and the coefficient is given The transfer function changes from 0.1 to 1 in 0.1 steps. The Bode diagram of Figure 2 As shown;

[0081] Step S303, fixed coefficient The value of remains unchanged, the coefficient The increase in the value of will cause the amplitude gain of the area to the left of the center frequency to be greater than 1, thereby introducing harmonics in the area near the center frequency. This problem can be avoided when ≤0.3; at the same time, it can be seen from the phase-frequency characteristics that the coefficient If the value of is too small, the phase change slope at the center frequency will be too large. Since the frequency convergence process is related to the cosine value of the phase, a phase change slope that is too large at the center frequency will reduce the convergence speed and increase the pulsation. Considering the rapidity and stability of convergence, the coefficient is selected. The value of is 0.3;

[0082] Step S304, fixed coefficient The value is 0.3, giving the coefficient The transfer function varies from 0.2 to 2 The Bode diagram of Figure 3 As shown, with the coefficient The value of decreases, its amplitude-frequency characteristic produces a severe peak in the center frequency area and the phase change slope at the center frequency also increases; as the coefficient As the value of increases, its amplitude-frequency and phase-frequency characteristics become smoother, which is beneficial to the suppression of harmonics and the rapid and smooth convergence of the resonant frequency estimation value; however, the coefficient If the value of is too large, the negative gain at low and high frequencies will decrease, which will weaken the harmonic suppression capability. Therefore, the coefficient is selected in a compromise. The value of is 1.4.

[0083] Step S4: Based on Step S1-Step S3, a fourth-order generalized integral resonance identifier in the continuous domain has been obtained. In order to realize digital application, the fourth-order generalized integral link is further discretized by using a pre-distortion discretization method to improve the error signal. The accuracy of

[0084] Specifically, the steps include:

[0085] In step S401, the pre-distortion discretization method used is expressed as:

[0086] (4)

[0087] In formula (4), To control the cycle;

[0088] Step S402: discretized fourth-order generalized integral link and its phase angle It is expressed as:

[0089]

[0090] = (4)

[0091] (5)

[0092] = (6)

[0093] In formula (6), for .

[0094] Step S5, extracting the error signal using a low-pass filter The DC component containing the resonance information The vibration frequency estimate of the speed error signal u is obtained through the integral controller ;

[0095] Specifically, the steps include:

[0096] Step S501: extract the error signal using a low-pass filter The DC component containing the resonance information , expressed as:

[0097] (7)

[0098] Step S502: Using an integral regulator to adjust the DC component Set it to zero, and we get:

[0099] (8)

[0100] Will Converge to , we can get the estimated value of the vibration frequency of u , estimated vibration frequency Used for resonance suppression of electric drive systems of deep-sea robots.

[0101] This embodiment is verified by experimental conditions:

[0102] (1) Experimental conditions: A 50 Hz sine wave is used as the signal for identification, and high-order harmonics and DC disturbances are injected in batches during the process;

[0103] Figure 4(a) is a waveform diagram of the traditional orthogonal principle method under different harmonic disturbances, and Figure 4(b) is a waveform diagram of the pre-distorted fourth-order generalized integral resonance identification method of this embodiment under different harmonic disturbances. As shown in Figure 4(a), after the injection of high-order harmonics and DC disturbances, the observed orthogonal components of the signal of the traditional orthogonal principle method begin to deviate from the correct value, resulting in an error in the final resonance identification result; as shown in Figure 4(b), the identification method proposed in this embodiment, because of its strong anti-disturbance ability, has almost no disturbance after the injection of high-order harmonics and DC disturbances, and can still accurately identify the result.

[0104] (2) Experimental conditions: The motor runs stably at 100 r / min. At this time, the motor has a resonance of about 33 Hz due to the installation of a flexible gear. The sampled speed error is used as a signal for identification, and the control frequencies of 10 k and 1 k are compared respectively.

[0105] Figure 5 The figure is a waveform comparison diagram of the traditional method (resonance identification method under traditional Euler discreteness) and the pre-distortion fourth-order generalized integral resonance identification method of this embodiment at different control frequencies. The upper waveform in the figure is the speed error signal used to identify the resonance, and the lower waveform is the resonance identification result of the two methods at a control frequency of 10k to 1k. Figure 5 As shown, under the control frequency of 10k, both the traditional method and the identification method proposed in this embodiment can accurately identify the resonance, but as the control frequency decreases from 10k to 1k, the traditional method has an identification error of about 10%, while the identification method proposed in this embodiment can still accurately identify the resonance due to its advantage of high identification accuracy.

[0106] (3) Experimental conditions: running time comparison in TMS320F28335;

[0107] Figure 6 This is a comparison chart of the running time of the traditional method 1 (traditional orthogonal principle method), the traditional method 2 (traditional second-order generalized integral resonance observation method) and the pre-distorted fourth-order generalized integral resonance identification method of this embodiment in TMS320F28335. Figure 6As shown, the identification method proposed in this embodiment has the advantage of small amount of calculation. Under the premise of strong anti-interference and high-precision online resonance identification performance, the identification method proposed in this embodiment only takes up 1.45μs more computing time than the traditional method, and the total computing time is only 3.66μs. Taking the 10k switching frequency as an example, it only takes up 3.66% of the terminal time.

[0108] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for identifying the fourth-order generalized integral resonance of the pre-distortion of the electric drive system of a deep-sea robot, characterized in that: The steps include: Step S1, sampling the motor speed and calculating the speed error signal u between the speed given value and the sampled value; Step S2, using a fourth-order generalized integral link to obtain a signal orthogonal to the speed error signal u , and multiply it with the speed error signal u to obtain the error signal ; Step S3, configuring resonance identification parameters based on high identification accuracy and strong harmonic suppression performance; Step S4, using the pre-distortion discretization method to discretize the fourth-order generalized integral link to improve the error signal The accuracy of Step S5, extracting the error signal using a low-pass filter The DC component containing the resonance information The vibration frequency estimate of the speed error signal u is obtained through the integral controller , estimated vibration frequency Used for resonance suppression of electric drive systems of deep-sea robots.

2. According to claim 1, a method for identifying fourth-order generalized integral resonance of a deep-sea robot electric drive system with pre-distortion, characterized in that: In step S2, the error signal Make judgments on low-order and high-order harmonic interference to facilitate subsequent resonance identification parameter configuration.

3. According to claim 2, a method for identifying fourth-order generalized integral resonance of a deep-sea robot electric drive system with pre-distortion, characterized in that: The judgment of low-order harmonic interference includes the following steps: Step S201, design a fourth-order generalized integral link: (1) In formula (1), is the transfer function of the fourth-order generalized integral link, j is the imaginary unit, is the angular frequency; and is the coefficient; is the estimated value of vibration frequency; is the amplitude of the fourth-order generalized integral link; is the phase angle of the fourth-order generalized integral link; Step S202: Based on formula (1), the amplitude of the low-order harmonics after conditioning by the fourth-order generalized integral link is ,because is an estimate of the vibration frequency, so Close to zero, which means the error signal There is no low-order harmonic interference.

4. According to claim 3, a method for identifying fourth-order generalized integral resonance of a deep-sea robot electric drive system with pre-distortion, characterized in that: The high-order harmonic interference judgment includes the following steps: Step S203, using a fourth-order generalized integral link to obtain a signal orthogonal to the speed error signal u It can be expressed as: (2) In formula (2), is the fundamental frequency amplitude of the fourth-order generalized integral link; is the fundamental frequency phase angle of the fourth-order generalized integral link; is the fundamental angular frequency of the fourth-order generalized integral link; is the k-th order harmonic amplitude of the fourth-order generalized integral link; is the kth order harmonic phase angle of the fourth order generalized integral link; is the kth order harmonic angular frequency of the fourth order generalized integral link; Step S204, based on formula (2), after the high-order harmonics are conditioned by the fourth-order generalized integral link, their amplitude decreases as the harmonic order increases, and their influence is negligible, which means that the error signal There is no high-order harmonic interference.

5. According to claim 4, a method for identifying fourth-order generalized integral resonance of pre-distortion of a deep-sea robot electric drive system is characterized in that: Error signal It is expressed as: (3) In formula (3), is the amplitude of the DC component in the speed error signal u; is the amplitude of the fundamental frequency component in the speed error signal u; is the amplitude of the higher harmonic components in the speed error signal u.

6. According to claim 2, a method for identifying fourth-order generalized integral resonance pre-distortion of a deep-sea robot electric drive system, characterized in that: In step S3, the configuration of the resonance identification parameters includes the following steps: Step S301, high recognition accuracy performance condition: transfer function The phase-frequency characteristic at the center frequency shows a phase lead of 90 degrees to ensure the convergence accuracy of the fourth-order generalized integral link to the orthogonal frequency; Strong harmonic suppression performance conditions: transfer function The amplitude gain at the center frequency is 1, and the amplitude at frequencies other than the center frequency is significantly attenuated to ensure that the fourth-order generalized integral link suppresses low-order and high-order harmonics; Step S302: Based on the conditions of high identification accuracy and strong harmonic suppression performance, set the coefficient The initial value is 1, and the coefficient is given The transfer function changes from 0.1 to 1 in 0.1 steps. Bode diagram; Step S303, fixed coefficient The value of remains unchanged, based on the rapidity and stability of the center frequency convergence, the coefficient The value of is adjusted to 0.3; Step S304, fixed coefficient The value is 0.3, giving the coefficient The transfer function varies from 0.2 to 2 Bode plot, based on the harmonic suppression capability, the coefficients The value of is adjusted to 1.

4.

7. According to claim 2, a method for identifying fourth-order generalized integral resonance of a deep-sea robot electric drive system with pre-distortion, characterized in that: Step S4 specifically includes the following steps: In step S401, the pre-distortion discretization method used is expressed as: (4) In formula (4), To control the cycle; Step S402: discretized fourth-order generalized integral link and its phase angle It is expressed as: = (4) (5) = (6) In formula (6), for .

8. According to claim 2, a method for identifying fourth-order generalized integral resonance of a deep-sea robot electric drive system with pre-distortion, characterized in that: Step S5 specifically includes the following steps: Step S501: extract the error signal using a low-pass filter The DC component containing the resonance information , expressed as: (7) Step S502: Using an integral regulator to adjust the DC component Set it to zero, and we get: (8) Will Converge to , we can get the estimated value of the vibration frequency of u .

Images