Self-adaptive augmentation controller design method based on frequency band response requirement

Abstract

The invention discloses a self-adaptive augmentation controller design method based on a frequency band response demand, and the method comprises the steps: reducing the nonlinearity of a conventional AAC controller, and obtaining an adjusted AAC controller; a sinusoidal signal is input into the AAC controller, and frequency domain response characteristics of the AAC controller are obtained through frequency sweep test analysis; the frequency domain of the AAC controller is correspondingly divided into three frequency bands, and parameter design requirements are proposed based on the response requirements of the AAC controller in different frequency bands; based on the determined parameter design requirements and priorities of different frequency band response requirements of the AAC, starting with a low-frequency characteristic signal with the lowest priority, designing an error term weight of an AAC controller, and introducing a high-frequency interference frequency domain response requirement to design a damping term weight; and based on the designed weight value and referring to the priority, performing coverage design on the insensitive section of the AAC controller. According to the method, the rapid design of each parameter of the self-adaptive augmentation controller is realized, and the purposes of simplicity, reliability, high efficiency and rapidness are achieved, so that the method has engineering practice application conditions.

Description

Technical Field

[0001] The invention relates to a design method of an adaptive augmented controller based on frequency band response requirements, and belongs to the technical field of spacecraft dynamics and control. Background Art

[0002] The present invention discloses a design method of an adaptive augmented controller based on frequency band response requirements. With large launch vehicles as the background, considering that launch vehicles are generally subject to two different types of interference, low frequency and high frequency, during flight, it is necessary to specially design the control parameters of the adaptive augmented controller so that it has the ability to cope with both low frequency and high frequency interference. At the same time, in view of the problems of many design parameters of the adaptive augmented controller, serious coupling between parameters, long parameter design cycle and high difficulty, the shortcomings of the prior art are:

[0003] There are few application cases of existing AAC controllers in the field of launch vehicles, and the academic community has not yet formed a systematic AAC controller parameter design method. Up to now, the technical design of AAC controller parameters generally adopts a closed-loop simulation trial and error method, or an autonomous optimization method based on an intelligent optimization algorithm. Among them, the closed-loop simulation trial and error method has a certain degree of engineering practicability, but the efficiency is too low, the workload is large, the work process is not systematic, and it is difficult to replicate. The method based on the intelligent optimization algorithm is too theoretical, has poor engineering practicability, and the algorithm is unstable, so the design effect is difficult to guarantee. Summary of the invention

[0004] The technical problem solved by the present invention is: to overcome the shortcomings of the prior art, to provide an adaptive augmented controller design method based on frequency band response requirements, to achieve rapid design of various parameters of the adaptive augmented controller through controller structure adjustment, frequency domain model acquisition, frequency band isolation and empirical formula design, to achieve the purpose of being simple, reliable, efficient and fast, and thus meet the conditions for engineering practice application.

[0005] The technical solution of the present invention is: a method for designing an adaptive augmentation controller based on frequency band response requirements, comprising:

[0006] Reducing the nonlinearity of a conventional adaptive augmented controller to obtain an adjusted AAC controller; the adaptive augmented controller is referred to as an adaptive augmented controller;

[0007] Input a sinusoidal signal to the AAC controller, and obtain the frequency domain response characteristics of the AAC controller through frequency sweep test analysis;

[0008] The frequency domain of the AAC controller is divided into three frequency bands: low frequency band, insensitive band and high frequency band; and parameter design requirements are proposed based on the response requirements of the AAC controller in different frequency bands;

[0009] Responsive design priorities for AAC controllers;

[0010] Based on the parameter design requirements and priorities of the AAC response requirements in different frequency bands, the error term weight of the AAC controller is designed starting from the low-frequency characteristic signal with the lowest priority, and the damping term weight is designed by introducing the high-frequency interference frequency domain response requirements;

[0011] Based on the designed weight value and referring to the priority, the insensitive segment of the AAC controller is covered and designed.

[0012] Preferably, the adjusted AAC controller formula is expressed as follows:

[0013]

[0014] Among them, k t is the final output coefficient, k t =k a +k0,k a is the integral part of the final output coefficient, and k0 is the constant part of the final output coefficient; is the rate of change of the final output coefficient, k tmax is the upper limit of the output coefficient, e r is the reference model deviation value, y s is the filter value of the spectrum damper; a is the weight of the error term, c is the weight of the damping term, and β is the weight of the regression term.

[0015] Preferably, the sinusoidal signal input to the AAC controller is:

[0016] e r =A·sin(ωt)

[0017] When the frequency ω value is close to 0 and approaches infinity, the adaptive final output coefficient of the AAC controller is solved as follows:

[0018]

[0019] The above results are analyzed by frequency sweep, and the frequency domain characteristics of the AAC controller only produce amplitude; this amplitude is equal to k in the low frequency band tmax , is equal to k0 in the high frequency band.

[0020] Preferably, the frequency domain response of the AAC controller is divided into three frequency bands:

[0021] Low frequency band, the AAC controller response is k tmax ;

[0022] Insensitive segment, the AAC controller responds slowly and transitions from the low frequency band to the high frequency band;

[0023] High frequency band, at this time the AAC controller responds to k0;.

[0024] Preferably, parameter design requirements are proposed based on the response requirements of AAC controllers in different frequency bands, specifically:

[0025] When the interference to the system is in the low frequency band or high frequency band, the design parameters should be able to quickly increase the AAC controller output k t Converges to k tmax or k0;

[0026] When the system is subject to interference in the insensitive range, the design parameters should try to make k t Oscillates around 1, i.e. does not respond to interference signals within this frequency band;

[0027] The design parameters that need to be considered include: low-frequency interference characteristic value e rc , error term weight a, high frequency interference characteristic value y sc , damping term weight c, high frequency interference characteristic value y s_max , low frequency interference limit value e r_max .

[0028] Preferably, the response design priority of the AAC controller is specifically:

[0029] The parameter design of the AAC controller gives priority to satisfying the high-frequency response of the AAC controller;

[0030] Secondly, when the interference is in the insensitive section, the adaptive final output coefficient k t Around 1;

[0031] Finally, AAC is able to respond correctly when interference is in the very low frequency band.

[0032] Preferably, the error term weight a is:

[0033]

[0034] Where: e rc For the maximum low-frequency interference, based on the task accuracy requirements, the maximum error amplitude acceptable to the system is taken; is the characteristic change rate; k a is the integral part of the final output coefficient, k tmax is the upper limit amplitude of the output coefficient;

[0035] Requirements r ≥e rc When the AAC controller converges to k at the characteristic rate of change tmax , e r is the reference model deviation value.

[0036] Preferably, the damping term weight c needs to satisfy the following constraints:

[0037]

[0038] Among them, y sc is the high-frequency interference characteristic value, k0 is the constant part of the final output coefficient, a is the error term weight, k tmax is the upper limit of the output coefficient, e rc The maximum low frequency interference.

[0039] Preferably, when designing the coverage of the insensitive segment of the AAC controller:

[0040] When the rigid body and elastic eigenvalues ​​of the AAC controller reach the limit at the same time in the insensitive section, the output change rate of the AAC controller is is 0, the high-frequency interference characteristic value y s_max It is expressed as:

[0041]

[0042] Among them, e r_max is the low-frequency disturbance limit value, k a is the integral part of the final output coefficient, k0 is the constant part of the final output coefficient, k tmax is the upper limit amplitude of the output coefficient, and c is the weight of the damping term.

[0043] Preferably, the regression term (k t -1) is of the order of 0.1; β(k t -1) is of the order of 0.01, and β is 0.1.

[0044] Compared with the prior art, the present invention has the following advantages:

[0045] (1) The present invention aims at the application scenario requirements of large launch vehicles, and the designed AAC controller parameters can cope with both low-frequency interference and high-frequency interference;

[0046] (2) After the present invention adjusts the structure of the AAC controller, its nonlinearity is reduced, so that the obtained frequency domain model has higher accuracy;

[0047] (3) By describing the response characteristics of the AAC controller in different frequency bands, the present invention can formulate corresponding control strategies for low-frequency and high-frequency interferences respectively, and the design process is free and flexible;

[0048] (4) The present invention carries out parameter design based on empirical formulas, and the design process is efficient and fast. Moreover, after one or two parameters are pre-selected, the remaining parameters can be calculated, and the design accuracy is higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the method steps of the present invention;

[0050] Figure 2 This is a typical frequency domain response curve of the AAC controller of the present invention. DETAILED DESCRIPTION

[0051] The present invention relates to a method for designing an adaptive augmentation control (AAC) controller based on frequency band response requirements, comprising the following steps:

[0052] 1) In view of the strong nonlinearity of the adaptive augmented controller, the controller formula is fine-tuned to reduce its nonlinearity as much as possible within the working range of the controlled object;

[0053] 2) Through frequency sweeping and mathematical derivation, the frequency domain response of the adaptive augmentation controller is roughly obtained. It is particularly noted that its response is mainly reflected in the amplitude and almost no phase is generated;

[0054] 3) Using the characteristics of its frequency domain amplitude response, the adaptive augmentation control working frequency band is divided into low frequency band, insensitive band and high frequency band, and the response requirements of the adaptive augmentation controller are designed according to different frequency bands;

[0055] 4) Starting from the control requirements of the controlled object itself, priorities are set for responses in different frequency bands. In particular, when the system is interfered by different frequency bands at the same time, the idea of ​​the worst operating condition indicator is introduced to determine the basis for the autonomous decision-making of the adaptive augmented controller;

[0056] 5) By pre-defining the motion range and dynamic performance of the controlled object and combining the frequency band response requirements, an analytical empirical formula is proposed to achieve rapid and accurate design of various parameters of the adaptive augmented controller.

[0057] The present invention proposes an adaptive augmentation controller (AAC) design method based on frequency band response requirements, and the specific implementation steps are as follows:

[0058] In step 1), the mathematical formula of the traditional AAC controller is as follows:

[0059]

[0060] k t =k a +k0 (2)

[0061] Among them, k t is the final output coefficient, k a is the integral part of the final output coefficient, is the rate of change of the final output coefficient, and k0 is the constant part of the final output coefficient. k a The derivative of k tmax is the upper limit of the output coefficient, e r is the reference model deviation value, y s is the filter value of the spectrum damper. a is the weight of the error term, c is the weight of the damping term, and β is the weight of the regression term. In the above formula, and -c·k a For nonlinear damping, k a The change in the integral is highly nonlinear and automatically converges to the pre-designed k t Limit value. In engineering applications, in order to facilitate the design of AAC controller parameters, the above formula needs to be fine-tuned. The adjusted formula is as follows:

[0062]

[0063] After the above adjustments, t When k is close to 1, a The integral shows good linear characteristics, which facilitates the subsequent parameter low-frequency interference characteristic value e rc , error term weight a, high frequency interference characteristic value y sc , damping term weight c, high frequency interference characteristic value y s_max , low frequency interference limit value e r_max design work.

[0064] In step 2), the frequency domain response characteristics of the AAC controller are obtained through frequency sweep test and formula derivation analysis. Assuming that the input of the AAC controller is a sinusoidal signal as shown below:

[0065] e r =A·sin(ωt) (4)

[0066] Based on the sinusoidal input signal described in formula (4), the damping loop signal can be written as:

[0067] y s ≈G dt ·A gt (e r ) 2

[0068] Where A is the signal amplitude, ω is the frequency, t is the time, G dt is the constant coefficient of the damping term, A gt For signal e r The filter gain after passing through the high-pass filter is not a constant value and is related to e rThe final output coefficient k of the AAC controller can be obtained by deriving the formula and assuming that the ω value is very small (ω≈0) and very large (ω→∞). t for:

[0069]

[0070] Among them, k t | ω≈0 and k t | ω→∞ They are the solutions of the AAC controller when the ω value is very small (ω≈0) and very large (ω→∞), respectively. After the above results are subjected to frequency sweep analysis, it is found that the frequency domain characteristics of the AAC controller produce almost no phase, only amplitude. This amplitude is equal to k in the low frequency band. tmax , is equal to k0 in the high frequency band. The frequency domain response curve of AAC controller is shown in Figure 2 shown.

[0071] In step 3), by analyzing the frequency domain response of the AAC controller, it is found that its frequency domain response can be divided into three frequency bands: low frequency band, at which the AAC controller response is k tmax ; Insensitive band, the AAC controller response slowly decreases, transitioning from the low frequency band to the high frequency band; High frequency band, the AAC controller response is k0. In view of this special property, two parameter design requirements are proposed: 1) When the interference to the system is in the low frequency band or high frequency band, the design parameters should be able to quickly increase the AAC controller output k t Converges to k tmax or k0; 2) When the interference to the system is in the insensitive section, the design parameters should try to make the final output coefficient k t Oscillates around 1, that is, does not respond to interference signals within this frequency band.

[0072] In step 4), after the parameter design requirements for different frequency bands are clarified in step 3), considering that in actual engineering applications, the system is generally subject to composite interference of multiple frequency bands, the response design priority of the AAC controller should be given. Taking the launch vehicle as an example, first, for this type of flexible structure, the parameter design of the AAC controller should be prioritized to meet the high-frequency response requirements; secondly, considering that the insensitive section of the AAC controller often coincides with the cutoff frequency of the flexible body, it should be ensured that when the interference is in the insensitive section, k t Around 1; finally, it ensures that AAC can respond correctly when interference is in the very low frequency band.

[0073] In step 5), based on the AAC different frequency band response requirements and priorities determined in the previous step, Figure 1 The design work was carried out according to the design process shown in Table 1. The low-frequency interference characteristic value e rc, error term weight a, high frequency interference characteristic value y sc , damping term weight c, high frequency interference characteristic value y s_max , low frequency interference limit value e r_max Design work:

[0074] i. Start with the low-frequency characteristic signal with the lowest priority, assuming that the system does not receive high-frequency disturbances at this time, and define the maximum low-frequency interference e rc and characteristic change rate Requirements r ≥e rc When the AAC controller converges to k at the characteristic rate of change tmax , e r is the reference model deviation value. rc Based on the task accuracy requirements, the maximum error amplitude acceptable to the system is taken. Therefore, the empirical formula for the error term weight is:

[0075]

[0076] ii. After confirming the first set of weights, introduce the high-frequency interference frequency domain response requirements, and define the high-frequency disturbance characteristic value y according to the priority rule sc According to y s and e r The relationship between y sc ≈G dt ·A gt (e rc ) 2 When the high-frequency disturbance reaches this level, even if the low-frequency disturbance is at its maximum level, rc , the rate of change of the final output coefficient of the AAC controller should be less than 0:

[0077]

[0078] So there is an empirical formula:

[0079]

[0080] Therefore, during the design process, c needs to satisfy the constraints of formula (9).

[0081] iii. After obtaining a and c using formulas (7) and (9), based on the designed weight values ​​and referring to the priority rules established in step 4, the design requirements of the insensitive section of the AAC controller are covered by selecting the low-frequency and high-frequency disturbance limit values. Since the AAC controller is required to have the rigid body and elastic eigenvalues ​​reach the limit at the same time in the insensitive section, the output change rate of the AAC controller is is 0. Therefore, we have:

[0082]

[0083] so:

[0084]

[0085] Among them, y s_max is the high-frequency disturbance eigenvalue, e r_max is the low-frequency disturbance limit value. Based on the above empirical formula, when any limit value is determined, the other one will also be determined accordingly.

[0086] Finally, considering the regression term (k t -1) is of the order of 0.1, while β(k t -1) should be kept in the range of 0.01, so β ​​is generally taken as 0.1. So far, the design process of all AAC controller parameters is completed.

[0087] Table 1 Design parameters and design methods of the present invention

[0088]

[0089] The contents not described in detail in the specification of the present invention belong to the prior art known to the professional and technical personnel in this field.

Claims

1. A design method for an adaptive augmented frequency controller based on frequency band response requirements, characterized in that include: The nonlinearity of the traditional adaptive augmented controller is reduced to obtain an adjusted adaptive augmented controller; The adaptive augmentation controller is denoted as AAC controller; Input a sinusoidal signal to the AAC controller, and obtain the frequency domain response characteristics of the AAC controller through frequency sweep test analysis; The frequency domain of the AAC controller is divided into three frequency bands: low frequency band, insensitive band and high frequency band; and parameter design requirements are proposed based on the response requirements of the AAC controller in different frequency bands; Responsive design priorities for AAC controllers; Based on the parameter design requirements and priorities of the AAC response requirements in different frequency bands, the error term weight of the AAC controller is designed starting from the low-frequency characteristic signal with the lowest priority, and the damping term weight is designed by introducing the high-frequency interference frequency domain response requirements; Based on the designed weight value and referring to the priority, the insensitive segment of the AAC controller is covered and designed.

2. The method for designing an adaptive augmentation controller based on frequency band response requirements according to claim 1, characterized in that: The adjusted AAC controller formula is expressed as follows: Among them, k t is the final output coefficient, k t =k a +k0,k a is the integral part of the final output coefficient, and k0 is the constant part of the final output coefficient; is the rate of change of the final output coefficient, k tmax is the upper limit of the output coefficient, e r is the reference model deviation value, y s is the filter value of the spectrum damper; a is the weight of the error term, c is the weight of the damping term, and β is the weight of the regression term.

3. The method for designing an adaptive augmentation controller based on frequency band response requirements according to claim 1, characterized in that: The sinusoidal signal input to the AAC controller is: e r =A sin(ωt) When the frequency ω value is close to 0 and approaches infinity, the adaptive final output coefficient of the AAC controller is solved as follows: The above results are analyzed by frequency sweep, and the frequency domain characteristics of the AAC controller only produce amplitude; this amplitude is equal to k in the low frequency band tmax , is equal to k0 in the high frequency band.

4. The method for designing an adaptive augmentation controller based on frequency band response requirements according to claim 1, characterized in that: The frequency domain response of the AAC controller is divided into three frequency bands: Low frequency band, the AAC controller response is k tmax ; Insensitive segment, the AAC controller responds slowly and transitions from the low frequency band to the high frequency band; High frequency band, at this time the AAC controller responds to k0;.

5. The method for designing an adaptive augmentation controller based on frequency band response requirements according to claim 4, characterized in that: Based on the response requirements of AAC controllers in different frequency bands, parameter design requirements are proposed, specifically: When the interference to the system is in the low frequency band or high frequency band, the design parameters should be able to quickly increase the AAC controller output k t Converges to k tmax or k0; When the system is subject to interference in the insensitive range, the design parameters should try to make k t Oscillates around 1, i.e. does not respond to interference signals within this frequency band; The design parameters that need to be considered include: low-frequency interference characteristic value e rc , error term weight a, high frequency interference characteristic value y sc , damping term weight c, high frequency interference characteristic value y s_max , low frequency interference limit value e r_max .

6. The method for designing an adaptive augmentation controller based on frequency band response requirements according to claim 1, characterized in that: The design priorities for the AAC controller response are: The parameter design of the AAC controller gives priority to satisfying the high-frequency response of the AAC controller; Secondly, when the interference is in the insensitive section, the adaptive final output coefficient k t Around 1; Finally, AAC is able to respond correctly when interference is in the very low frequency band.

7. The method for designing an adaptive augmentation controller based on frequency band response requirements according to claim 1, characterized in that: The error term weight a is: Where: e rc For the maximum low-frequency interference, based on the task accuracy requirements, the maximum error amplitude acceptable to the system is taken; is the characteristic change rate; k a is the integral part of the final output coefficient, k tmax is the upper limit amplitude of the output coefficient; Requirements r ≥e rc When the AAC controller converges to k at the characteristic rate of change tmax , e r is the reference model deviation value.

8. The method for designing an adaptive augmentation controller based on frequency band response requirements according to claim 1, characterized in that: The constraints that the damping term weight c needs to satisfy are: Among them, y sc is the high-frequency interference characteristic value, k0 is the constant part of the final output coefficient, a is the error term weight, k tmax is the upper limit of the output coefficient, e rc The maximum low frequency interference.

9. The method for designing an adaptive augmentation controller based on frequency band response requirements according to claim 1, characterized in that: When overwriting the insensitive segment of the AAC controller: When the rigid body and elastic eigenvalues ​​of the AAC controller reach the limit at the same time in the insensitive section, the output change rate of the AAC controller is is 0, the high-frequency interference characteristic value y s_max It is expressed as: Among them, e r_max is the low-frequency disturbance limit value, k a is the integral part of the final output coefficient, k0 is the constant part of the final output coefficient, k tmax is the upper limit amplitude of the output coefficient, and c is the weight of the damping term.

10. The method for designing an adaptive augmentation controller based on frequency band response requirements according to claim 2, characterized in that: The regression term of the AAC controller (k t -1) is of the order of 0.1; β(k t -1) is of the order of 0.01, and β is 0.1.

Images