A performance evaluation method for aerospace electromechanical servo system

Abstract

The present invention discloses a performance evaluation method for an aerospace electromechanical servo system, which solves the problems in the forward development of aerospace electromechanical servo systems, such as the large number of performance evaluation index parameters and the small number of available data set samples, which make it difficult to quantify performance indicators and evaluate product solution performance. At the same time, it also provides a simple and efficient performance quantitative evaluation method for parameter setting, performance degradation analysis, fault monitoring, etc. of aerospace electromechanical servo systems.

Description

Technical Field

[0001] The present invention belongs to the technical field of electromechanical servo system design and simulation, and in particular relates to a performance evaluation method for an aerospace electromechanical servo system. Background Art

[0002] The existing performance evaluation methods for aerospace electromechanical servo systems need to assess multiple dimensions of indicators, including but not limited to: overshoot, rise time, steady-state accuracy, transition time, maximum speed, etc. in transient characteristics; nominal position gain, maximum deflection angle, zero position deviation, position loop width in position characteristics; and amplitude and phase angle corresponding to each frequency point in frequency characteristics.

[0003] With the growing thirst for knowledge in the vast unknowns of space under the new circumstances, the pace of innovation in spacecraft research and development is accelerating. The aerospace servo system is a key executive subsystem in the control systems of launch vehicles, such as launch vehicles or space planes. It is primarily responsible for driving loads such as the engine (or engine nozzle), air rudders, or deformation mechanisms to predetermined positions to control the vehicle's attitude. Its performance directly determines the success or failure of the vehicle's flight. As the most powerful, energy-intensive, and environmentally challenging system in launch vehicles and space planes, apart from the engine, with the most complex dynamic characteristics, the aerospace servo system is also a key component of its electrification process. Performance evaluation of the electromechanical servo system involves complex coupled analysis across multiple disciplines, including mechanics, energy management, electromagnetics, power electronics, and control. However, existing evaluation methods suffer from numerous indicators and testing methods that are difficult to apply in the early stages of forward development. This makes it difficult to quantify the performance of aerospace electromechanical servo systems in the early stages of design. Product finalization relies on multiple rounds of physical optimization iterations, and the iteration rate lags behind the pace of spacecraft development, a situation that is difficult to reverse. Therefore, in order to achieve accurate estimation of design performance in the initial stage of forward R&D of aerospace electromechanical servo systems with fewer indicator constraints, it is necessary to provide a set of simple and efficient performance evaluation methods with low sample dependence, which can provide a new way to improve the efficiency of forward R&D of aerospace electromechanical servo systems. Summary of the Invention

[0004] The technical problem solved by the present invention is to overcome the shortcomings of the existing technology and provide a performance evaluation method for aerospace electromechanical servo systems, which solves the problems in the forward development of aerospace electromechanical servo systems caused by the large number of performance evaluation index parameters and the small number of available data set samples, resulting in difficulty in quantifying performance indicators and evaluating product solution performance.

[0005] The purpose of the present invention is achieved through the following technical solutions: A method for evaluating the performance of an aerospace electromechanical servo system, comprising:

[0006] Acquire data sets for aerospace electromechanical servo system performance evaluation, including position command signals, transmission mechanism linear displacement, motor mechanical speed, and motor quadrature-axis current;

[0007] According to the collected linear displacement of the transmission mechanism, the linear displacement of the transmission mechanism is converted into the load angular displacement using geometric trigonometric relationship;

[0008] According to the collected position command signal and the converted load angular displacement, the angular displacement curve fitting degree is calculated and used as a quantitative evaluation index of the position followability of the aerospace electromechanical servo system;

[0009] According to the collected motor mechanical speed, the Rayleigh entropy of the three-band motor mechanical speed is calculated using short-time Fourier transform and frequency-division Rayleigh entropy. The maximum value of the three-band Rayleigh entropy is used as a quantitative evaluation index of the speed stability of the aerospace electromechanical servo system.

[0010] Based on the collected motor quadrature-axis current, the Rayleigh entropy of the three-band motor quadrature-axis current is calculated using short-time Fourier transform and frequency-division Rayleigh entropy. The per-unit resonant peak-to-mean value is obtained through fast Fourier transform and threshold screening. The maximum value of the three-band Rayleigh entropy and the per-unit resonant peak-to-mean value are used as quantitative evaluation indicators for the output power smoothness and anti-interference performance of the aerospace electromechanical servo system.

[0011] Based on the obtained quantitative evaluation indicators of position followability, speed stability, output power smoothness and anti-interference performance, the performance level evaluation of aerospace electromechanical servo system is given.

[0012] Preferably, the data set used for the performance evaluation of aerospace electromechanical servo systems is a data set obtained under non-specific or fixed instructions. The data sample sizes of the position command signal and the transmission mechanism linear displacement must be consistent. The data of the transmission mechanism linear displacement, the motor mechanical speed and the motor quadrature-axis current must be data samples within the same time period, and the data sample sizes may be inconsistent.

[0013] Preferably, the geometric trigonometric relationship is:

[0014]

[0015] Among them, θ is the load angular displacement, r is the rocker arm length, l is the linear displacement of the transmission mechanism, and b is the initial zero position length of the transmission mechanism.

[0016] Preferably, the calculating curve fitting degree includes:

[0017] The command signal is represented by a sinusoidal signal, a step signal or any combination of the two. For the angular displacement obtained under the sinusoidal signal command, the lag phase angle degree must be obtained by the single-variable single-target gradient descent method first, and the determination coefficient of the load angular displacement and the position command signal after compensating for the lag phase angle is used as the curve fitting degree. If it is a step signal, the determination coefficient of the load angular displacement and the position command signal is directly used as the curve fitting degree. If it is any combination of the two, the corresponding determination coefficients are calculated for the sinusoidal signal and the step signal in the combination respectively, and the smallest determination coefficient is selected as the curve fitting degree of the command signal.

[0018] Preferably, the single variable refers to the variable of the load angular displacement and the lag phase angle of the position command signal, and its value range is between [-2π, 0] rad; the single target refers to the determination coefficient of the load angular displacement data set and the position command signal data set after adding the lag phase angle compensation.

[0019] Preferably, the coefficient of determination calculation formula is:

[0020]

[0021] Where: R 2 is the coefficient of determination; k is the sample index in the data set; θ k * is the kth sample of the position command signal data set; θ k is the kth sample of the load angular displacement data set; N is the total number of samples in the data set.

[0022] Preferably, frequency division means ignoring the influence of frequencies above 500 Hz, dividing the frequency domain into three sections: below 30 Hz, 30 Hz-150 Hz, and 150 Hz-500 Hz, which are represented by very low frequency, low frequency, and mid-low frequency respectively.

[0023] Preferably, the normalized resonance peak average value is to take 10% of the fundamental wave amplitude as a threshold value and take the average of the amplitudes of the frequency points whose amplitudes are greater than 10% of the fundamental wave amplitude for normalization processing.

[0024]

[0025] Where, is the normalized peak-to-average value of the resonance, M is the total number of frequency points whose amplitude exceeds 10% of the fundamental amplitude, j is the index value of each frequency point, P f0 is the fundamental frequency amplitude.

[0026] Preferably, the performance level evaluation of the aerospace electromechanical servo system is given according to the obtained quantitative evaluation index of position followability, quantitative evaluation index of speed stability, quantitative evaluation index of output power smoothness and anti-interference performance, and by comparing with the performance quantitative evaluation table.

[0027] Preferably, the performance quantitative evaluation table is:

[0028]

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] (1) This invention significantly reduces the dimensions of traditional aerospace electromechanical servo system performance evaluation indicators and the reliance of the performance evaluation process on instructions and data sample sizes. It can support the performance estimation of aerospace electromechanical servo system solutions in the early stages of the forward R&D process when there are no refined indicator requirements.

[0031] (2) The present invention significantly reduces the dimensions of the performance evaluation indicators of the aerospace electromechanical servo system, and reduces the compromise difficulty and labor cost of debugging and adjusting the control parameters of the aerospace electromechanical servo system;

[0032] (3) The performance evaluation formula of the present invention is simple, has a small amount of calculation, and has good embedded development portability, and can be used for quantitative analysis and evaluation of online performance degradation and fault monitoring of aerospace electromechanical servo systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0034] Figure 1 This is a flow chart of a method for evaluating the performance of an aerospace electromechanical servo system according to the present invention;

[0035] Figure 2 The hardware composition, control architecture and spatial geometric layout diagram of the aerospace electromechanical servo system involved in the present invention;

[0036] Figure 3 It is a time domain curve fitting diagram of the position command signal and the load angular displacement involved in the present invention;

[0037] Figure 4 A short-time Fourier transform time-frequency characteristic diagram of the mechanical speed of the motor involved in the present invention;

[0038] Figure 5 The short-time Fourier transform time-frequency characteristic diagram of the quadrature-axis current of the motor involved in the present invention;

[0039] Figure 6 A fast Fourier transform frequency characteristic diagram of the quadrature axis current of the motor involved in the present invention; DETAILED DESCRIPTION

[0040] The following will refer to the attached Figure 1-6 Exemplary embodiments of the present disclosure are described in more detail. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, unless there is a conflict, the embodiments of the present disclosure and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0041] like Figure 1 As shown, the specific steps of the aerospace electromechanical servo system performance evaluation method of the present invention are as follows:

[0042] First, data is collected from the signal generator, linear displacement sensor, speed sensor, and current sensor respectively. If it is a product test, the measured data is converted into a sample set with a unified time starting point and the corresponding real physical quantity at that time point through the tester, including the position command signal θ * r , transmission mechanism linear displacement l r , motor mechanical speed n r and the motor quadrature axis current i qr (The subscript r indicates a real object); if it is a digital simulation of the servo scheme, it is converted into a sample set with a unified time starting point and the corresponding simulation physical quantity at that time point through software with certain data processing functions such as Excel or Matlab, including the position command signal θ * v , transmission mechanism linear displacement l v , motor mechanical speed n v and the motor quadrature axis current i qv (The subscript v indicates a virtual model).

[0043] The aerospace electromechanical servo system is a key executive subsystem in the control system of an aircraft represented by a launch vehicle or a space plane. It is mainly responsible for driving the engine (or engine nozzle) or air rudder or deformation mechanism and other loads to the predetermined position to control the attitude of the aircraft. Its performance directly determines the success or failure of the aircraft's navigation. The hardware components of the aerospace electromechanical servo system include: servo power supply, control driver, motor, reducer, and transmission mechanism. The aerospace electromechanical servo system usually adopts a three-loop closed-loop control architecture of position loop, speed loop, and current loop. Usually, the motor in the aerospace electromechanical servo system adopts the direct axis current (i d ) is equal to zero. The data set used for the performance evaluation of aerospace electromechanical servo system includes the position command signal θ *, transmission mechanism linear displacement l, motor mechanical speed n and motor quadrature axis current i q The data set used for the performance evaluation of aerospace electromechanical servo systems can be a data set obtained under specific or fixed instructions, and the position instruction signal θ * The data sample size of the transmission mechanism linear displacement l and the transmission mechanism linear displacement l, the motor mechanical speed n and the motor quadrature axis current i q The data sample sizes of the three may be inconsistent, but they must be data samples from the same time period.

[0044] Second, if Figure 2 As shown, the transmission mechanism linear displacement l is converted into the load angular displacement θ using geometric trigonometric relationships. The present invention uses geometric trigonometric relationships to indirectly convert the transmission mechanism linear displacement into the load angular displacement rather than directly acquiring the load angular displacement because aerospace vehicles fly at high speeds, generating high temperatures and high vibrations that are unsuitable for, and no sensors are suitable for, such harsh conditions. The geometric trigonometric conversion relationship between the transmission mechanism linear displacement and the load angular displacement can be expressed as:

[0045]

[0046] Among them, θ is the load angular displacement, r is the rocker arm length, l is the linear displacement of the transmission mechanism, and b is the initial zero position length of the transmission mechanism.

[0047] Third, the position command signal θ * and load angular displacement θ are plotted as Figure 3 Angular displacement time domain curve, using the determination coefficient to calculate the load angular displacement θ and position command signal θ * The degree of fitting is used as a quantitative evaluation index of the position following performance of aerospace electromechanical servo system.

[0048] The command signal can be obtained by a sinusoidal signal, a step signal, or any combination of the two. For the angular displacement obtained under the sinusoidal signal command, the lag phase angle between the load angular displacement and the position command signal is used as a variable, with a value range of [-2π, 0] rad. The determination coefficient between the load angular displacement and the position command signal after compensating for the lag phase angle is used as the target. The gradient descent method is used to obtain the lag phase angle degree, and then the determination coefficient between the load angular displacement and the position command signal after compensating for the lag phase angle is used as the curve fit. If it is a step signal, the determination coefficient between the load angular displacement and the position command signal is directly used as the curve fit. If it is any combination of a sinusoidal signal and a step signal, the corresponding determination coefficients are calculated for the sinusoidal signal and the step signal in the combination according to the above method, and the smallest determination coefficient is selected as the curve fit of the command signal.

[0049] The coefficient of determination is calculated as follows:

[0050]

[0051] Where: R 2 is the coefficient of determination; k is the sample index in the data set; θ k * is the kth sample of the position command signal data set; θ k is the kth sample of the load angular displacement data set; N is the total number of samples in the data set.

[0052] Fourth, perform short-time Fourier transform on the motor mechanical speed n and draw Figure 4 The short-time Fourier transform time-frequency characteristic diagram of the motor mechanical speed is shown. Based on the short-time Fourier transform results, the influence of frequencies above 500 Hz is ignored. The Rayleigh entropy calculation formula is used to obtain the Rayleigh entropy of the motor mechanical speed n in three frequency bands: extremely low frequency below 30 Hz, low frequency between 30 Hz and 150 Hz, and medium and low frequency between 150 Hz and 500 Hz. The maximum value of the Rayleigh entropy of the motor mechanical speed n in the three frequency bands is used as the quantitative evaluation index of the speed stability of the aerospace electromechanical servo system.

[0053] The relevant formula for short-time Fourier transform is:

[0054]

[0055] Where S(t,ω) is a two-dimensional function defined in time and frequency by discrete samples; N is the total number of samples in the data set; x(m) is the input signal, ω(m) is the window function, ω(mt) indicates the offset of the window function by t samples, and i is the imaginary unit.

[0056] Fifth, the motor quadrature axis current i q Perform short-time Fourier transform and draw Figure 5 The short-time Fourier transform time-frequency characteristic diagram of the motor quadrature axis current is shown in Figure 2. q Perform fast Fourier transform and draw Figure 6 The fast Fourier transform frequency characteristic diagram of the motor quadrature axis current is based on the short-time Fourier transform results, ignoring the influence of frequencies above 500Hz. The Rayleigh entropy calculation formula is used to obtain the motor quadrature axis current i in three frequency bands: very low frequency below 30Hz, low frequency 30Hz-150Hz, and medium-low frequency 150Hz-500Hz. q Rayleigh entropy; Based on the fast Fourier transform results, 10% of the fundamental amplitude is used as the threshold, and the average amplitude of each frequency point with an amplitude greater than 10% of the fundamental amplitude is taken for standardization. q The maximum value of Rayleigh entropy and the normalized resonant peak-to-mean value are used as quantitative evaluation indicators for the smoothness and anti-interference performance of the output power of aerospace electromechanical servo systems.

[0057] The relevant formula for fast Fourier transform is:

[0058] F[k]=DFT(e[j])+e (-2πki / N) DFT(o[j])

[0059]

[0060]

[0061] Where j and k are the sample indices in the dataset; x j is the jth data sample, e[j] is x 2j , is an even sequence of samples; o[j] is x 2j+1 , is an odd sequence of samples; N represents the total number of samples in the data set; i is an imaginary unit.

[0062] The relevant formula for calculating the resonance peak-to-average value is:

[0063]

[0064] Where, is the normalized peak-to-average value of the resonance, M is the total number of frequency points whose amplitude exceeds 10% of the fundamental amplitude, j is the index value of each frequency point, P f0 is the fundamental frequency amplitude.

[0065] The formula for calculating Rayleigh entropy is:

[0066]

[0067] Where: ɑ is the order, R P ɑ is the ɑ-order Rayleigh entropy, and P(t,ω) is the data set representing the time-frequency distribution.

[0068] Sixth, the quantitative evaluation indicators of position followability, speed stability, output power smoothness and anti-interference performance are referred to the performance quantitative evaluation table in Table 1 to give the performance level evaluation of the aerospace electromechanical servo system as excellent, good, qualified and unqualified.

[0069] The performance quantitative evaluation table must at least include indicators such as position tracking, speed stability, output power smoothness, and anti-interference performance. However, targeted additions and adjustments to indicator magnitudes may be made based on the power level of the AESC servo system, the current health status of the product, the control strategy used, and the characteristics of the driven load. The performance level of the AESC servo system can be divided into four levels: excellent, good, qualified, and unqualified.

[0070] Table 1 Performance quantitative evaluation table

[0071]

[0072] Examples

[0073] Take the direct-drive aerospace electromechanical servo system (without reducer) as an example, in which the load is an air rudder, the motor is a three-phase permanent magnet synchronous motor, the transmission mechanism is a planetary ball screw, and the control architecture is a three-loop closed-loop control of position loop, speed loop, and current loop. The position loop adopts PID control strategy, the speed loop adopts PI control, and the current loop adopts PI control. The motor adopts the direct axis current (i d ) equal to zero; the position command signal is a sinusoidal command signal. To maximize the beneficial effects of the present invention, two data sources are presented in this example: a direct-drive aerospace electromechanical servo system product and a corresponding virtual mapping performance analysis model.

[0074] (1) Install the direct-drive aerospace electromechanical servo system product on the air rudder load platform, complete the strong and weak electrical connections, tester debugging and interface inspection, and collect data from the signal generator, linear displacement sensor, speed sensor, and current sensor respectively, and obtain a sample set with a unified time starting point containing a time point and the real physical quantity corresponding to the time point, including the position command signal θ * r , transmission mechanism linear displacement l r , motor mechanical speed n r and the motor quadrature axis current i qr (The subscript r indicates a real object); based on the Matlab Simulink platform, a direct-drive aerospace electromechanical servo system performance simulation analysis model is simultaneously built, and the model parameter assignment and initialization are completed based on the actual state attributes of the product (the control parameters and state parameters are completely consistent, but the performance of the model and the actual object will definitely be different and cannot be completely consistent, which does not affect the effect of the present invention). A sample set with a unified time starting point containing a time point and the corresponding simulation physical quantity at that time point is obtained, including the position command signal θ * v , transmission mechanism linear displacement l v , motor mechanical speed n v and the motor quadrature axis current i qv (Subscript v indicates a virtual model). In this example, the position command signal θ * r and θ * v They are all sinusoidal signals with an amplitude of 1° and a frequency of 6 rad / s, that is, θ * r =θ * v=1*sin(6t), assume that both the object and the model run 3 sine cycles, a total of 3.14s, with the same sampling period, and finally obtain the measured position command signal θ with the same time starting point and 2400 data. * r , model position command signal θ * v , measured transmission mechanism linear displacement l r , linear displacement of the model transmission mechanism l v , Measure the motor mechanical speed n r , Model motor mechanical speed n v , measured motor quadrature axis current i qr , model motor quadrature axis current i qv .

[0075] (2) By Figure 2 The spatial layout of the physical transmission mechanism collected in step (1) can be known by using the geometric triangulation relationship. r and the linear displacement l of the model transmission mechanism v Converted into actual load angular displacement θ r and the model load angular displacement θ v , where the geometric trigonometric relationship is as follows:

[0076]

[0077] Where θ is the load angular displacement; r is the rocker arm length, and both the measured value and the model value in this example are 0.1 m; l is the linear displacement of the transmission mechanism; and b is the initial zero-position length of the transmission mechanism, and both the measured value and the model value in this example are 0.4 m.

[0078] (3) Based on the measured load angular displacement θ r and the model load angular displacement θ v With the position command signal θ * r and θ * v The lag phase angle is used as a variable, and the determination coefficient of the load angular displacement and the position command signal after compensating the lag phase angle is used as the target. The measured load angular displacement θ is obtained by the gradient descent method. r With the position command signal θ * r The phase lag is 0.0552rad, and the model load angular displacement θ v With the position command signal θ * v The phase lag is 0.0533 rad, and the measured load angular displacement θ obtained in step (1) is r and the model load angular displacement θ v and the position command signal θ after compensating the lag phase angle *rc and θ * vc Draw Figure 3 Angular displacement time domain curve. Calculate the measured load angular displacement θ using the coefficient of determination r The measured position command signal θ after compensating the lag phase angle * rc The fitting degree is 0.9971, and the model load angular displacement θ is calculated using the determination coefficient v and the model position command signal θ after compensating the lag phase angle * vc The fitting degree is 0.9958, and the fitting degree is used as a quantitative evaluation index of the position following performance of aerospace electromechanical servo system.

[0079] The calculation formula of the coefficient of determination involved in step (3) is as follows:

[0080]

[0081] Where R 2 is the coefficient of determination; k is the sample index in the data set; θ k * is the kth sample of the position command signal data set; θ k is the kth sample of the load angular displacement data set; N is the total number of samples in the data set, and in this example, N=2400.

[0082] (4) The measured motor mechanical speed n obtained in step (1) r and the model motor mechanical speed n v Perform short-time Fourier transform and draw Figure 4 The short-time Fourier transform time-frequency characteristic diagram of the motor mechanical speed. Based on the short-time Fourier transform results, ignoring the influence of frequencies above 500Hz, the Rayleigh entropy calculation formula is used to obtain the measured motor mechanical speed n in three frequency bands: very low frequency below 30Hz, low frequency between 30Hz and 150Hz, and medium and low frequency between 150Hz and 500Hz. r and the model motor mechanical speed n v The Rayleigh entropy of the three frequency bands of the measured motor mechanical speed from low to high is [5.07, 7.74, 8.70], and the Rayleigh entropy of the three frequency bands of the model motor mechanical speed from low to high is [4.97, 7.58, 7.85], and the maximum value of the Rayleigh entropy of the three frequency bands of the motor mechanical speed (the measured maximum value is 8.70, and the model maximum value is 7.85) is used as the quantitative evaluation index of the speed stability of the aerospace electromechanical servo system.

[0083] The relevant formula for the short-time Fourier transform involved in step (4) is:

[0084]

[0085] Where S(t,ω) is a two-dimensional function defined by discrete samples in time and frequency; N is the total number of samples in the data set, in this example N = 2400; x(m) is the input signal, which is the measured motor mechanical speed n. r and the model motor mechanical speed n v ω(m) is the window function. In this example, the frequency axis length is set to 512. ω(mt) indicates that the window function has an offset of t samples. In this example, the Hanning window function is selected for specific window function characterization. Its formula is In this example, the Hanning window length is set to 128, and i is an imaginary unit.

[0086] Frequency division means ignoring the influence of frequencies above 500Hz and dividing the frequency domain into three segments: below 30Hz, 30Hz-150Hz, and 150Hz-500Hz, which are represented by very low frequency, low frequency, and mid-low frequency respectively.

[0087] The Rayleigh entropy calculation formula involved in step (4) is:

[0088]

[0089] Where: ɑ is the order. In this example, ɑ=3 in the specific calculation process, so R P ɑ In this example, it is the third-order Rayleigh entropy, and P(t,ω) is the data set representing the time-frequency distribution.

[0090] (5) The measured motor quadrature axis current i obtained in step (1) qr and the quadrature-axis current i of the model motor qv Perform short-time Fourier transform and draw Figure 5 The short-time Fourier transform time-frequency characteristic diagram of the motor quadrature axis current. Based on the short-time Fourier transform results, ignoring the influence of frequencies above 500Hz, the Rayleigh entropy calculation formula is used to obtain the measured motor quadrature axis current i in three frequency bands: very low frequency below 30Hz, low frequency between 30Hz and 150Hz, and medium and low frequency between 150Hz and 500Hz. qr and the quadrature-axis current i of the model motor qv The Rayleigh entropy of the three frequency bands of the measured motor quadrature-axis current from low to high is [5.37, 7.97, 8.82], and the Rayleigh entropy of the three frequency bands of the model motor quadrature-axis current from low to high is [5.17, 7.78, 8.86]. The description and expression of short-time Fourier transform, frequency division and Rayleigh entropy are the same as those in step (4) and will not be repeated here.

[0091] The measured motor quadrature axis current i qr and the quadrature-axis current i of the model motor qv Perform fast Fourier transform and draw Figure 6Fast Fourier transform frequency characteristics of the motor's quadrature-axis current.

[0092] The relevant formula for the fast Fourier transform involved in step (5) is:

[0093] F[k]=DFT(e[j])+e (-2πki / N) DFT(o[j])

[0094]

[0095]

[0096] Where j and k are the sample indices in the dataset; x j is the jth data sample, e[j] is x 2j , is an even sequence of samples; o[j] is x 2j+1 , is an odd sequence of samples; N represents the total number of samples in the data set, that is, N = 2400; i is an imaginary unit.

[0097] Based on the fast Fourier transform results, the measured motor quadrature axis current i qr and the quadrature-axis current i of the model motor qv 10% of the fundamental frequency amplitude is used as the threshold value, and the mean value of the amplitude of each frequency point with an amplitude greater than 10% of the fundamental amplitude is obtained for standardization. Here, there are 9 frequency points with an amplitude greater than 10% of the fundamental amplitude, and the measured motor quadrature axis current i is obtained. qr The normalized resonant peak-to-average value is 10.04, and the quadrature-axis current i of the model motor is obtained. qv The per-unit resonance peak average value is 10.51.

[0098] The maximum value of the Rayleigh entropy in the three frequency bands of the motor quadrature-axis current (the maximum value of the measured result is 8.82, and the maximum value of the model result is 8.86) and the per-unit resonance peak-to-mean value (the per-unit resonance peak-to-mean value of the measured result is 10.04, and the per-unit resonance peak-to-mean value of the model result is 10.51) are used as quantitative evaluation indicators for the output power smoothness and anti-interference performance of the aerospace electromechanical servo system.

[0099] The relevant formula for the resonance peak-to-average value involved in step (5) is:

[0100]

[0101] Where, is the normalized peak-to-average value of the resonance, M is the total number of frequency points whose amplitude exceeds 10% of the fundamental amplitude, j is the index value of each frequency point, P f0 is the fundamental frequency amplitude.

[0102] (6) By referring to the quantitative evaluation index of position followability, the quantitative evaluation index of speed stability, and the quantitative evaluation index of output power smoothness and anti-interference performance in Table 1, it can be concluded that the aerospace electromechanical servo system product used for example description has the following characteristics: r The measured position command signal θ after compensating the lag phase angle * r The fitting degree is 0.9971, so its position tracking performance is "good"; the measured motor mechanical speed n r The maximum value of the three-band Rayleigh entropy is 8.70, so its speed stability is "qualified"; the measured motor quadrature axis current i qr The maximum value of the three-band Rayleigh entropy is 8.82, and the per-unit resonance peak average is 10.04, so its output power smoothness and anti-interference performance are "good".

[0103] The simulation model of the aerospace electromechanical servo system used for example description is v and the model position command signal θ after compensating the lag phase angle * v The fitting degree is 0.9958, so its position tracking performance is "good"; the mechanical speed of the model motor n v The maximum value of the three-band Rayleigh entropy is 7.85, so its speed stability is "good"; the quadrature axis current i qv The maximum value of the three-band Rayleigh entropy is 8.86, and the per-unit resonance peak average is 10.51, so its output power smoothness and anti-interference performance are "good".

[0104] It should be noted that the quantitative evaluation criteria of the evaluation table given in Table 1 are only used to illustrate the practical effects of this example. In actual applications, the performance quantitative evaluation table must at least include indicators such as position tracking, speed stability, output power smoothness, and anti-interference. However, it can also be targeted to add and adjust the magnitude of evaluation indicators according to the different power levels of the aerospace electromechanical servo system, the current health status of the product, the control strategy adopted, and the characteristics of the driven load object.

[0105] Parts of the present invention that are not described in detail belong to the common knowledge of those skilled in the art.

Claims

1. A method for evaluating the performance of an aerospace electromechanical servo system, characterized in that: include: Acquire data sets for aerospace electromechanical servo system performance evaluation, including position command signals, transmission mechanism linear displacement, motor mechanical speed, and motor quadrature-axis current; According to the collected linear displacement of the transmission mechanism, the linear displacement of the transmission mechanism is converted into the load angular displacement using geometric trigonometric relationship; According to the collected position command signal and the converted load angular displacement, the angular displacement curve fitting degree is calculated and used as a quantitative evaluation index of the position followability of the aerospace electromechanical servo system; According to the collected motor mechanical speed, the Rayleigh entropy of the three-band motor mechanical speed is calculated using short-time Fourier transform and frequency-division Rayleigh entropy. The maximum value of the three-band Rayleigh entropy is used as a quantitative evaluation index of the speed stability of the aerospace electromechanical servo system. Based on the collected motor quadrature-axis current, the Rayleigh entropy of the three-band motor quadrature-axis current is calculated using short-time Fourier transform and frequency-division Rayleigh entropy. The per-unit resonant peak-to-mean value is obtained through fast Fourier transform. The maximum value of the three-band Rayleigh entropy and the per-unit resonant peak-to-mean value are used as quantitative evaluation indicators for the output power smoothness and anti-interference performance of the aerospace electromechanical servo system. Based on the obtained quantitative evaluation indicators of position followability, speed stability, output power smoothness and anti-interference performance, the performance level evaluation of aerospace electromechanical servo system is given.

2. The aerospace electromechanical servo system performance evaluation method according to claim 1, characterized in that: The data set used for the performance evaluation of aerospace electromechanical servo systems is a data set obtained under non-specific or fixed instructions. However, the data sample sizes of the position command signal and the transmission mechanism linear displacement must be consistent. The data of the transmission mechanism linear displacement, the motor mechanical speed, and the motor quadrature-axis current must be data samples within the same time period.

3. The aerospace electromechanical servo system performance evaluation method according to claim 1, characterized in that: The geometric trigonometric relationship is: Among them, θ is the load angular displacement, r is the rocker arm length, l is the linear displacement of the transmission mechanism, and b is the initial zero position length of the transmission mechanism.

4. The aerospace electromechanical servo system performance evaluation method according to claim 1, characterized in that: The calculation of the curve fitting degree includes: The command signal is represented by a sinusoidal signal, a step signal or any combination of the two. For the angular displacement obtained under the sinusoidal signal command, the lag phase angle degree must be obtained by the single-variable single-target gradient descent method first, and the determination coefficient of the load angular displacement and the position command signal after compensating for the lag phase angle is used as the curve fitting degree. If it is a step signal, the determination coefficient of the load angular displacement and the position command signal is directly used as the curve fitting degree. If it is any combination of the two, the corresponding determination coefficients are calculated for the sinusoidal signal and the step signal in the combination respectively, and the smallest determination coefficient is selected as the curve fitting degree of the command signal.

5. The aerospace electromechanical servo system performance evaluation method according to claim 4, characterized in that: The single variable refers to the lag phase angle between the load angular displacement and the position command signal, and its value range is between [-2π, 0] rad; the single target refers to the determination coefficient between the load angular displacement data set and the position command signal data set after adding the lag phase angle compensation.

6. The aerospace electromechanical servo system performance evaluation method according to claim 4, characterized in that: The formula for calculating the coefficient of determination is: Where: R 2 is the coefficient of determination; k is the sample index in the data set; θ k * is the kth sample of the position command signal data set; θ k is the kth sample of the load angular displacement data set; N is the total number of samples in the data set.

7. The aerospace electromechanical servo system performance evaluation method according to claim 1, characterized in that: Frequency division means ignoring the influence of frequencies above 500Hz and dividing the frequency domain into three segments: below 30Hz, 30Hz-150Hz, and 150Hz-500Hz, which are represented by very low frequency, low frequency, and mid-low frequency respectively.

8. The aerospace electromechanical servo system performance evaluation method according to claim 1, characterized in that: The normalized resonant peak average value is to take 10% of the fundamental wave amplitude as a threshold value and take the average of the amplitudes of the frequency points whose amplitudes are greater than 10% of the fundamental wave amplitude for normalization processing.

9. The aerospace electromechanical servo system performance evaluation method according to claim 1, characterized in that: According to the quantitative evaluation indexes of position followability, speed stability, output power smoothness and anti-interference performance, and by comparing with the performance quantitative evaluation table, the performance level evaluation of the aerospace electromechanical servo system is given.

10. The aerospace electromechanical servo system performance evaluation method according to claim 9, characterized in that: The performance quantitative evaluation table is:

Images