A High-Frequency Fatigue Machine Iterative Learning Control Method Based on Robust Disturbance Observer

Abstract

This invention discloses an iterative learning control method for a high-frequency fatigue machine based on a robust disturbance observer. The method includes: obtaining a mathematical model of the high-frequency fatigue machine through system identification using a sinusoidal frequency sweep method; designing a robust disturbance observer based on the mathematical model of the high-frequency fatigue machine; estimating system disturbances using the robust disturbance observer based on the control input and output of the iterative controller, thereby achieving cancellation with the actual system disturbances; ultimately improving the disturbance rejection capability and tracking performance of the high-frequency fatigue machine. This invention not only fully utilizes iterative learning control to achieve high tracking accuracy but also utilizes a robust disturbance observer to improve system control performance.

Description

Technical Field

[0001] This invention relates to the technical field of fatigue testing machines, and more particularly to a high-frequency fatigue machine iterative learning control method based on a robust disturbance observer. Background Technology

[0002] Fatigue performance is a crucial parameter in engineering design, directly impacting product reliability and long-term stability. As the execution platform for fatigue testing, fatigue testing machines must not only meet stability requirements but also achieve high technical standards in dynamic response and tracking performance. Linear motors, due to their advantages of high speed, high precision, and low maintenance, are increasingly being used as the drive source for high-frequency fatigue testing machines. However, the lack of intermediate buffers in linear motors makes them extremely sensitive to end effects, friction, load disturbances, and parameter perturbations, increasing the difficulty of control. Therefore, to improve system control performance, it is essential to design a control method with high tracking accuracy and strong robustness, which has become a pressing technical problem to be solved. Summary of the Invention

[0003] To address the problems in the existing technology, the present invention aims to overcome the shortcomings of the existing technology and provide a high-frequency fatigue machine iterative learning control method based on a robust disturbance observer, which can improve the tracking accuracy and dynamic response of the high-frequency fatigue machine and improve the system control performance by utilizing a robust disturbance observer.

[0004] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:

[0005] A high-frequency fatigue machine iterative learning control method based on a robust disturbance observer includes the following steps:

[0006] Step 1: Obtain the mathematical model of the high-frequency fatigue machine by performing system identification using the sinusoidal frequency sweep method;

[0007] Step 2: Employ a robust disturbance observer based on a mathematical model of a high-frequency fatigue machine;

[0008] Step 3: Select a suitable iterative learning law and design an iterative learning controller;

[0009] Step 4: Based on the control input and output of the iterative controller, the system disturbance is estimated through a robust disturbance observer to achieve cancellation with the actual system disturbance;

[0010] Step 5: Based on the current tracking error of the sub-high frequency fatigue machine and the control output and tracking error of the previous iteration, calculate the next control output through the iterative learning controller;

[0011] Step 6: Control the high-frequency fatigue machine by iteratively learning the output of the controller, obtain the output data and tracking error of the high-frequency fatigue machine for the next iteration, and record the control output of this iteration.

[0012] Step 7: Repeat steps 4-6 until the fatigue test is completed.

[0013] Preferably, in step 1, the high-frequency fatigue machine system includes a system identification module, a disturbance observer, an iterative learning controller, a position sensor module, and a high-frequency fatigue machine mechanical device; wherein, the system identification module is used to obtain the transfer function of the high-frequency fatigue machine;

[0014] The robust disturbance observer is used to estimate the disturbance of the system and then compensate for it equivalently.

[0015] The controller receives control commands from the system and uses the data obtained from the sensor modules to achieve precise control of the fatigue testing machine through iterative learning.

[0016] The position sensor module is used to collect the position information of the high-frequency fatigue machine as feedback input to the controller;

[0017] The high-frequency fatigue machine mechanical device, as the control object of the system, uses a linear motor as its power source.

[0018] Preferably, in step 1, the system identification method is to apply a sinusoidal sweep frequency excitation signal to the input end of the high-frequency fatigue machine, the position sensor module collects the output position data, and the transfer function of the system is obtained by the least squares method.

[0019] Preferably, in step 2, the robust disturbance observer is calculated as follows:

[0020]

[0021] It is an estimate of the system disturbance, P. n (s) is the system model obtained by the sinusoidal frequency sweep method. It is the inverse model of the system, y k (s) is the output of the system at the kth iteration, u k+1 Q(s) is the control output of the system in the (k+1)th iteration, and Q(s) and Q1(s) are the filters to be designed. The filters are designed based on the complex frequency domain, and s is the complex frequency.

[0022] More preferably, Q(s) includes a low-pass filter Q1(s) and a notch filter Q1(s), and Q(s) is designed as follows:

[0023] Q(s) = 1 - (1 - Q1(s))·Q2(s)

[0024] in, The filter is designed based on the complex frequency domain, where s is the complex frequency, T1 is the time constant of the low-pass filter, and ξ1 is the damping coefficient of the low-pass filter; T NF1,2 These are the time constants of the notch filter, ξ and ξ. NF1,2 These are the damping coefficients of the notch filter; the function of the low-pass filter Q1(s) is to make the system physically implementable while suppressing high-frequency noise. The function of the notch filter Q2(s) is to maintain strong interference suppression capability when the low-pass filter's ξ1 is small, and to improve system stability.

[0025] Preferably, in step 3, the iterative learning controller calculates the control output:

[0026] u k+1 (s)=C(s)e k+1 (s)+Q D (s)(v k (s)+L(s)e k (s))

[0027] L(s) is the learning filter, Q D (s) is a low-pass filter, e k+1 (s) is the position error signal in the (k+1)th iteration, v k (s) is both the ILC output signal of the k-th iteration and the ILC input signal of the (k+1)-th iteration, u k+1 C(s) is the control signal for the (k+1)th iteration, and C(s) is the coefficients of k... p and k d The PD feedback controller, where s is the complex frequency.

[0028] More preferably, in order to retain the system's effective control input signal and suppress high-frequency random disturbance signals, a low-pass filter Q is designed. D The cutoff frequency ω of (s) n The low-pass filter Q is greater than the natural frequency of the desired trajectory and less than the frequency of the random disturbance signal. D The form of (s) is:

[0029]

[0030] PD-type ILCs are suitable for position servo systems and offer excellent control performance. The learning filter L(s) is selected as follows:

[0031] L(s) = k1 + k2s

[0032] Where k1 is the proportional coefficient, k2 is the differential coefficient, the filter is designed based on the complex frequency domain, and s is the complex frequency.

[0033] Preferably, based on the control input and output of the iterative controller, the system disturbance is estimated by a robust disturbance observer to achieve cancellation with the actual system disturbance.

[0034] Preferably, the next control output is calculated by an iterative learning controller based on the tracking error of the current sub-high frequency fatigue machine and the control output and tracking error of the previous iteration.

[0035] Preferably, the high-frequency fatigue machine is controlled by the output of the iterative learning controller to obtain the output data and tracking error of the high-frequency fatigue machine in the next iteration, while the control output of this iterative learning is recorded.

[0036] Preferably, the controller runs iteratively multiple times until the fatigue test ends.

[0037] Compared with the prior art, the present invention has the following obvious and prominent substantive features and significant advantages:

[0038] 1. This invention improves the tracking accuracy of high-frequency fatigue machines while shortening the iteration cycle and setup time; moreover, this invention has a faster response speed and stronger anti-interference capability, solving the problem of poor anti-interference of linear motors as power sources for high-frequency fatigue machines.

[0039] 2. This invention not only improves the tracking accuracy and dynamic response of high-frequency fatigue testing machines, but also enhances system control performance by utilizing a robust disturbance observer. Attached Figure Description

[0040] Figure 1 This is a block diagram illustrating the principle of a high-frequency fatigue machine iterative learning control method based on a robust disturbance observer, according to a preferred embodiment of the present invention.

[0041] Figure 2 The flowchart shows a preferred embodiment of the high-frequency fatigue machine iterative learning control method based on a robust disturbance observer according to the present invention.

[0042] Figure 3 The figure shows the simulation results of the high-frequency fatigue machine iterative learning control method based on a robust disturbance observer according to Embodiment 3 of the present invention. Detailed Implementation

[0043] The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It is obvious that the described embodiments are only a part of, and not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention.

[0044] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0045] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0046] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0047] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0048] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0049] The above solution will be further described below with reference to specific embodiments. The preferred embodiments of the present invention are described in detail below:

[0050] Example 1:

[0051] In this embodiment, see Figure 1 A high-frequency fatigue machine iterative learning control method based on a robust disturbance observer includes the following steps.

[0052] Step 1: Obtain the mathematical model of the high-frequency fatigue machine by performing system identification using the sinusoidal frequency sweep method;

[0053] Step 2: Employ a robust disturbance observer based on a mathematical model of a high-frequency fatigue machine;

[0054] Step 3: Select a suitable iterative learning law and design an iterative learning controller;

[0055] Step 4: Based on the control input and output of the iterative controller, the system disturbance is estimated through a robust disturbance observer to achieve cancellation with the actual system disturbance;

[0056] Step 5: Based on the current tracking error of the sub-high frequency fatigue machine and the control output and tracking error of the previous iteration, calculate the next control output through the iterative learning controller;

[0057] Step 6: Control the high-frequency fatigue machine by iteratively learning the output of the controller, obtain the output data and tracking error of the high-frequency fatigue machine for the next iteration, and record the control output of this iteration.

[0058] Step 7: Repeat steps 4-6 until the fatigue test is completed.

[0059] In this embodiment, the high-frequency fatigue machine iterative learning control method based on a robust disturbance observer can improve the tracking accuracy and dynamic response of the high-frequency fatigue machine, and improve the system control performance by utilizing the robust disturbance observer.

[0060] Example 2:

[0061] This embodiment is basically the same as Embodiment 1, except that:

[0062] In this embodiment, as Figure 1 and Figure 2 In step 1, the high-frequency fatigue machine system includes a system identification module, a disturbance observer, an iterative learning controller, a position sensor module, and a high-frequency fatigue machine mechanical device; wherein, the system identification module is used to obtain the transfer function of the high-frequency fatigue machine;

[0063] The robust disturbance observer is used to estimate the disturbance of the system and then compensate for it equivalently.

[0064] The controller receives control commands from the system and uses the data obtained from the sensor modules to achieve precise control of the fatigue testing machine through iterative learning.

[0065] The position sensor module is used to collect the position information of the high-frequency fatigue machine as feedback input to the controller;

[0066] The high-frequency fatigue machine mechanical device, as the control object of the system, uses a linear motor as its power source.

[0067] In this embodiment, in step 1, the system identification method is to apply a sinusoidal sweep frequency excitation signal to the input end of the high-frequency fatigue machine, the position sensor module collects the output position data, and the transfer function of the system is obtained by the least squares method.

[0068] In this embodiment, the robust disturbance observer is calculated in step 2 as follows:

[0069]

[0070] It is an estimate of the system disturbance, P. n (s) is the system model obtained by the sinusoidal frequency sweep method. It is the inverse model of the system, y k (s) is the output of the system at the kth iteration, u k+1 Q(s) is the control output of the system in the (k+1)th iteration, and Q(s) and Q1(s) are the filters to be designed. The filters are designed based on the complex frequency domain, and s is the complex frequency.

[0071] In this embodiment, in step 3, the iterative learning controller calculates the control output:

[0072] u k+1 (s)=C(s)e k+1 (s)+Q D (s)(v k (s)+L(s)e k (s))

[0073] L(s) is the learning filter, Q D (s) is a low-pass filter, e k+1 (s) is the position error signal in the (k+1)th iteration, v k (s) is both the ILC output signal of the k-th iteration and the ILC input signal of the (k+1)-th iteration, u k+1 C(s) is the control signal for the (k+1)th iteration, and C(s) is the coefficients of k... p and k d The PD feedback controller, where s is the complex frequency.

[0074] This embodiment improves the tracking accuracy of the high-frequency fatigue machine while shortening the iteration cycle and setup time; moreover, the present invention has a faster response speed and stronger anti-interference capability, solving the problem of poor anti-interference of linear motors as power sources for high-frequency fatigue machines.

[0075] Example 3:

[0076] This embodiment is basically the same as the above embodiments, except that:

[0077] In this embodiment, a high-frequency fatigue machine iterative learning control method based on a robust disturbance observer includes:

[0078] Reference Figure 1 , Figure 1 This is a block diagram of the iterative learning control principle of a high-frequency fatigue machine based on a robust disturbance observer. It consists of five parts: an iterative learning controller, a system identification module, a robust disturbance observer, a high-frequency fatigue machine, and a position sensor module.

[0079] Reference Figure 2 , Figure 2 The flowchart for high-frequency fatigue machine iterative learning control based on a robust disturbance observer includes the following steps:

[0080] S1: By applying a sinusoidal sweep frequency excitation signal w to the input terminal of the high-frequency fatigue machine k The position sensor module collects and outputs position data y k The transfer function of the system is obtained by using the least squares method.

[0081] S2: Design a robust perturbation observer:

[0082]

[0083] It is an estimate of the system disturbance, P. n (s) is the system model obtained by sinusoidal frequency sweep. It is the inverse model of the system, y k (s) is the output of the system at the kth iteration, u k+1 Q(s) is the control output of the system in the (k+1)th iteration, and Q(s) and Q1(s) are the filters to be designed. The filters are designed based on the complex frequency domain, and s is the complex frequency.

[0084] In this embodiment, Q(s) includes a low-pass filter Q1(s) and a notch filter Q1(s), and Q(s) is designed as follows:

[0085] Q(s) = 1 - (1 - Q1(s))·Q2(s)

[0086] in, The filter is designed based on the complex frequency domain, where s is the complex frequency, T1 is the time constant of the low-pass filter, and ξ1 is the damping coefficient of the low-pass filter; T NF1,2 These are the time constants of the notch filter, ξ and ξ. NF1,2 These are the damping coefficients of the notch filter; the function of the low-pass filter Q1(s) is to make the system physically implementable while suppressing high-frequency noise. The function of the notch filter Q2(s) is to maintain strong interference suppression capability when the low-pass filter's ξ1 is small, and to improve system stability.

[0087] S3: High-frequency fatigue machine iterative learning control based on robust disturbance observer, wherein: the iterative learning controller:

[0088] u k+1 (s)=C(s)e k+1 (s)+Q D (s)(v k (s)+L(s)e k (s))

[0089] L(s) is the learning filter, Q D (s) is a low-pass filter, e k+1 (s) is the position error signal in the (k+1)th iteration, v k (s) is both the ILC output signal of the k-th iteration and the ILC input signal of the (k+1)-th iteration, u k+1 C(s) is the control signal for the (k+1)th iteration, and C(s) is the coefficients of k... p and k d The PD feedback controller, where s is the complex frequency.

[0090] In this embodiment, in order to retain the system's effective control input signal and suppress high-frequency random disturbance signals, a low-pass filter Q is designed. D The cutoff frequency ω of (s) n The frequency is greater than the natural frequency of the desired trajectory, but less than the frequency of the random disturbance signal. Low-pass filter Q D The form of (s) is:

[0091]

[0092] PD-type ILCs are suitable for position servo systems and offer excellent control performance. The learning filter L(s) is selected as follows:

[0093] L(s) = k1 + k2s

[0094] Where k1 is the proportional coefficient, k2 is the differential coefficient, the filter is designed based on the complex frequency domain, and s is the complex frequency.

[0095] S4: Based on the control input and output of the iterative controller, the system disturbance is estimated through a robust disturbance observer to achieve cancellation with the actual system disturbance;

[0096] S5: Based on the tracking error of the current sub-high frequency fatigue machine and the control output and tracking error of the previous iteration, the control output for the next iteration is calculated through the iterative learning controller;

[0097] S6: Control the high-frequency fatigue machine by iteratively learning the output of the controller, obtain the output data and tracking error of the high-frequency fatigue machine for the next iteration, and record the control output of this iterative learning.

[0098] S7: Repeat steps 4-6 until the fatigue test is completed.

[0099] In this embodiment, in order to verify and illustrate the technical effects of the method used in this embodiment, this embodiment selects traditional iterative learning control and compares it with this method for testing. The experimental results are compared using scientific demonstration methods to verify the real effect of this method.

[0100] The method designed in this invention is verified using the following simulation examples:

[0101] Traditional high-frequency fatigue machine iterative learning control is poor at suppressing non-repetitive disturbances. Non-repetitive disturbances will weaken the tracking effect of iterative learning control and may even cause iterative learning control to diverge.

[0102] This embodiment can introduce a robust disturbance observer while keeping the system control performance unchanged, thereby improving the system control accuracy and stability issues caused by non-repetitive disturbances.

[0103] A simulation platform was built using MATLAB / SIMULINK, and non-repeatable noise was added to the system. Figure 3 It can be seen that the high-frequency fatigue machine iterative learning control method based on a robust disturbance observer in this embodiment can effectively reduce the impact of non-repetitive disturbances while maintaining the high steady-state performance of the system. This embodiment's high-frequency fatigue machine iterative learning control method, based on a robust disturbance observer, estimates system disturbances based on the control input and output of the iterative controller, achieving cancellation with the actual system disturbances; ultimately improving the disturbance rejection capability and tracking performance of the high-frequency fatigue machine. This embodiment not only fully utilizes iterative learning control to achieve high tracking accuracy but also utilizes a robust disturbance observer to improve system control performance.

[0104] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above embodiments. Various changes can be made according to the purpose of the invention. Any changes, modifications, substitutions, combinations or simplifications made based on the spirit and principle of the technical solution of the present invention shall be equivalent substitutions. As long as they meet the purpose of the invention and do not deviate from the technical principle and inventive concept of the present invention, they shall fall within the protection scope of the present invention.

Claims

1. A high-frequency fatigue machine iterative learning control method based on a robust disturbance observer, characterized in that: Includes the following steps, Step 1: Obtain the mathematical model of the high-frequency fatigue machine by performing system identification using the sinusoidal frequency sweep method; Step 2: Employ a robust disturbance observer based on a mathematical model of a high-frequency fatigue machine; Step 3: Select a suitable iterative learning law and design an iterative learning controller; Step 4: Based on the control input and output of the iterative controller, the system disturbance is estimated through a robust disturbance observer to achieve cancellation with the actual system disturbance; Step 5: Based on the current tracking error of the sub-high frequency fatigue machine and the control output and tracking error of the previous iteration, calculate the next control output through the iterative learning controller; Step 6: Control the high-frequency fatigue machine by iteratively learning the output of the controller, obtain the output data and tracking error of the high-frequency fatigue machine for the next iteration, and record the control output of this iteration. Step 7: Repeat steps 4-6 until the fatigue test is completed; In step 2, the robust disturbance observer is calculated as follows: It is an estimate of the system disturbance, P. n (s) is the system model obtained by the sinusoidal frequency sweep method. It is the inverse model of the system, y k (s) is the output of the system at the kth iteration, u k+1 Q(s) is the control output of the system in the (k+1)th iteration, Q(s) and Q1(s) are the filters to be designed. The filters are designed based on the complex frequency domain, and s is the complex frequency. Q(s) includes a low-pass filter Q1(s) and a notch filter Q2(s), and Q(s) is designed as follows: Q(s) = 1 - (1 - Q1(s))·Q2(s) in, The filter is designed based on the complex frequency domain, where s is the complex frequency, T1 is the time constant of the low-pass filter, and ξ1 is the damping coefficient of the low-pass filter; TNF 1，2 These are the time constants of the notch filter, ξ and ξ. NF1，2 These are the damping coefficients of the notch filter.

2. The high-frequency fatigue machine iterative learning control method based on a robust disturbance observer according to claim 1, characterized in that: In step 1, the high-frequency fatigue machine system includes a system identification module, a disturbance observer, an iterative learning controller, a position sensor module, and a high-frequency fatigue machine mechanical device; wherein, the system identification module is used to obtain the transfer function of the high-frequency fatigue machine; The robust disturbance observer is used to estimate the disturbance of the system and then compensate for it equivalently. The controller receives control commands from the system and uses the data obtained from the sensor modules to achieve precise control of the fatigue testing machine through iterative learning. The position sensor module is used to collect the position information of the high-frequency fatigue machine as feedback input to the controller; The high-frequency fatigue machine mechanical device, as the control object of the system, uses a linear motor as its power source.

3. The high-frequency fatigue machine iterative learning control method based on a robust disturbance observer according to claim 2, characterized in that: In step 1, the system identification method is to apply a sinusoidal sweep frequency excitation signal to the input of the high-frequency fatigue machine, the position sensor module collects the output position data, and the transfer function of the system is obtained by the least squares method.

4. The high-frequency fatigue machine iterative learning control method based on a robust disturbance observer according to claim 1, characterized in that: In step 3, the iterative learning controller calculates the control output: u k+1 (s)=C(s)e k+1 (s)+Q D (s)(v k (s)+L(s)e k (s)) L(s) is the learning filter, Q D (s) is a low-pass filter, e k+1 (s) is the position error signal in the (k+1)th iteration, v k (s) is both the ILC output signal of the k-th iteration and the ILC input signal of the (k+1)-th iteration, u k+1 C(s) is the control signal for the (k+1)th iteration, and C(s) is the coefficients of k... p and k d The PD feedback controller, where s is the complex frequency.

5. The high-frequency fatigue machine iterative learning control method based on a robust disturbance observer according to claim 4, characterized in that: In step 3, a low-pass filter Q is designed. D The cutoff frequency ωn of (s) is greater than the natural frequency of the desired trajectory and less than the frequency of the random disturbance signal. The low-pass filter Q D The form of (s) is: Using P D Type ILC is suitable for position servo systems, and the learning filter L(s) is selected: L(s) = k1 + k2s Where k1 is the proportional coefficient, k2 is the differential coefficient, the filter is designed based on the complex frequency domain, and s is the complex frequency.

Images