A fuzzy-reasoning-based active self-adaptive offload control method for launch vehicle

Abstract

The application discloses a kind of based on fuzzy inference's launch vehicle active adaptive load shedding control method, for the launch vehicle big wind area load shedding problem under uncertain wind field, research launch vehicle intelligent adaptive active load shedding control method, realize big wind area load index requirement, the main content of the application includes launch vehicle active load shedding control scheme design, bandwidth parameter adaptive change's extended state observer design, parameter intelligent adaptive tuning strategy design based on fuzzy inference.The application of a kind of based on fuzzy inference's launch vehicle active adaptive load shedding control method, from the angle of active load shedding and robustness&adaptability&intelligentization promotion, research launch vehicle intelligent adaptive active load shedding control method, for improving the load shedding capacity of our future launch vehicle and its international competitiveness has important theoretical significance.

Description

Technical Field

[0001] This invention belongs to the field of rocket control, specifically relating to an active adaptive load reduction control method for launch vehicles based on fuzzy reasoning. Background Technology

[0002] Launch vehicle load reduction technology employs guidance and control techniques to mitigate the impact of high-altitude winds on launch vehicle flight safety, thereby reducing the aerodynamic load and ensuring safe and reliable flight in high-altitude wind zones. For existing rockets, the primary purpose of load reduction is to increase launch probability; however, for newly developed rockets, if load reduction control technology can be effectively utilized to reduce the aerodynamic load, the structural strength requirements can be appropriately lowered during the design phase, reducing the structural mass and thus increasing the launch vehicle's carrying capacity and reducing launch costs.

[0003] Traditional load reduction methods typically employ ballistic correction compensation, incorporating pre-measured high-altitude wind information into the control system to compensate for the wind load's impact on the launch vehicle. This load reduction compensation method is simple in principle, relatively easy to implement, and highly reliable, while not affecting the original control system structure; it is often referred to as a passive load reduction method. However, the accuracy of pre-launch high-altitude wind compensation depends on the precision of the pre-programmed high-altitude wind information, and its ability to suppress real-time uncertainties in wind load is poor. Therefore, it cannot fully compensate for the impact of the wind field on the load during actual flight, lacking robustness and adaptability to real-time uncertain wind fields.

[0004] Therefore, traditional passive load reduction methods face an urgent need for comprehensive improvement in design flexibility, load reduction initiative, and system intelligence. Summary of the Invention

[0005] This invention proposes an active adaptive load reduction control method for launch vehicles based on fuzzy reasoning. It studies the intelligent adaptive active load reduction control method for launch vehicles from the perspectives of active load reduction and the improvement of robustness, adaptability and intelligence.

[0006] A fuzzy reasoning-based active adaptive load reduction control method for launch vehicles includes the following steps: (1) Designing an active load reduction control scheme for the launch vehicle, including a combination of rocket body attitude feedback and overload feedback. The traditional load reduction control equation is as follows:

[0007]

[0008] In the formula, This indicates the pitch attitude angle deviation under the arrow system; A represents the attitude angle feedback gain coefficient; y1 This indicates the apparent acceleration signal (rocket body coordinate system) output by the normal acceleration meter; Φ Ay1(z) represents the attitude angle and overload feedback correction network; K P K I This represents the ratio and integral coefficient of the overload feedback.

[0009] (3) Design an extended state observer with adaptive bandwidth parameter variation, including extended state estimation and compensation methods;

[0010] ①The basic extended state observer is as follows:

[0011]

[0012] In the formula, y = ω z1 Indicates system output, ω z1 z1 and z2 represent the extended state observer's estimates of the system state and uncertain dynamics, respectively; b0 represents the control efficiency coefficient estimate; u represents the total control output of the controller; β1 > 0, β2 > 0 represent the observer coefficients, typically β1 = 2β n ,β2=β n 2 , where β n >0 indicates the observer's basic bandwidth parameter.

[0013] ② Construct an adaptive extended state observer, employing a linear state observer design method with adaptively varying bandwidth parameter β. n Adaptation is expressed as a nonlinear function in the following form:

[0014]

[0015] Where, β n0 >0 indicates the bandwidth limit value; κ>0 indicates the growth rate coefficient. The larger the value, the faster the curve rises, which means it is closer to the traditional design scheme.

[0016] ③ The traditional extended state observer will be improved into the following adaptive parametric form:

[0017]

[0018] With the introduction of an adaptive extended state observer, the original active load shedding control algorithm will be improved to the following form:

[0019]

[0020] (3) Design a parameter intelligent adaptive tuning strategy based on fuzzy inference:

[0021] A. Determine the compensation amount for the extended state observer and introduce adaptive allocation coefficients. The active adaptive load reduction control algorithm is as follows:

[0022]

[0023] B. Load margin Q α_YD As the input signal to the fuzzy controller, the adaptive parameters will be... As the output signal of the fuzzy controller, the load margin is divided into four fuzzy regions: PB (positive large), PM (positive medium), PS (positive small), and ZO (zero value). It is divided into three fuzzy regions: PB, PM and PS.

[0024] C. Design fuzzy mapping rules, including 4 fuzzy IF-THEN mapping rules, with the following rule format:

[0025] R (j) :IF Q α_YD is A (j) THEN K IGC is B (j)

[0026] Among them, A (j) and B (j) Let Q represent the input and output fuzzy sets, respectively. The input and output fuzzy sets of the fuzzy controller are described using trigonometric functions, where the input variable Q... α_YD and output variables The domain ranges are [0.0, 1.0) and [Ukmin, Ukmax], respectively; Ukmin < 1.0 and Ukmax > 1.0 are the minimum and maximum values ​​of the domain of the output variable, respectively.

[0027] D. The centroid method is used for defuzzification. The adaptive coefficients are obtained through a fuzzy logic reasoning mechanism, and the calculation method is as follows:

[0028]

[0029] Preferably, the load margin expression in step (3) is defined as follows:

[0030]

[0031] Among them, Q α_max Q represents the aerodynamic load index value. α =q|α| represents the current aerodynamic load value, q represents the dynamic pressure, and α represents the angle of attack. When the calculated value is negative, Q is... α_YD Set it to 0. Therefore, when Q... α Stay away from Q α_max When the load margin is large, when Q α Approaching Q α_max When Q decreases, the load margin value decreases. αGreater than or equal to Q α_max At that time, the load margin becomes 0.

[0032] Preferably, the method for dividing the fuzzy region in step (3) includes: using fuzzy inference to... The value is adjusted online adaptively when the load margin Q α_YD If too small, increase the size appropriately. The value achieves a further load reduction effect; when the load margin Q α_YD When the value is large, reduce it appropriately. This value is used to free up more control capabilities for attitude angle deviation control.

[0033] A mathematical simulation verification and comparison experiment includes the above-mentioned active adaptive load reduction control method for launch vehicles. The specific implementation method includes: under typical deviation combination states, comparing the basic load reduction control scheme, the load reduction control scheme with an extended state observer, and the adaptive load reduction control method of the present invention. The maximum values ​​of the corresponding aerodynamic loads under the three load reduction control schemes are 228185.831, 188831.839, and 144136.100, respectively.

[0034] The present invention provides an active adaptive load reduction control method for launch vehicles based on fuzzy reasoning. Addressing the urgent needs of launch vehicle development, compared to the traditional load reduction control method combining rocket attitude feedback and overload feedback, this invention significantly reduces aerodynamic loads in high-wind areas, achieving a better load reduction effect and meeting the load design requirements for high-wind areas. Specific innovations are as follows:

[0035] (1) The active adaptive load reduction control method proposed in this invention can, on the one hand, observe the uncertain disturbances of the system in real time and make control compensation, and on the other hand, it can adaptively adjust the key control parameters according to the current load margin. Therefore, the method has stronger robustness and adaptability, and thus presents better load reduction capability.

[0036] (2) This invention proposes the concept of load margin, and based on this, it further proposes a fuzzy reasoning method for control allocation coefficients based on the current load margin, which can integrate human experience and knowledge into the parameter tuning strategy design process without relying on precise mapping relationship, thereby improving the intelligence of the load reduction control system.

[0037] (3) The method of the present invention fully reflects the characteristics of practical engineering design. It makes robustness, adaptability and intelligent improvement on the basis of traditional solutions, which is convenient for engineering application. At the same time, it has high method versatility and can be extended to other types of flight objects with load reduction requirements. Attached Figure Description

[0038] Figure 1 This is a comparison curve of aerodynamic loads under the three control schemes in Example 2. Detailed Implementation

[0039] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection claimed by the present invention.

[0040] Example 1

[0041] This application discloses an active adaptive load reduction control method for launch vehicles based on fuzzy reasoning, comprising the following steps:

[0042] (1) Design an active load reduction control scheme for launch vehicles

[0043] To achieve active load reduction in high-wind areas, most studies employ overload feedback in their load reduction control schemes. Compared to angle-of-attack control, this approach allows for more direct load control and enhances the load reduction effect. Therefore, traditional active load reduction control schemes often combine rocket attitude feedback and overload feedback. Attitude angle deviation feedback ensures the rocket tracks the programmed angle, while overload feedback ensures timely adjustments under heavy loads. The traditional load reduction control equations are as follows:

[0044]

[0045] In the formula,

[0046] —Pitch angle deviation under the arrow system;

[0047] —Attitude angle feedback gain coefficient;

[0048] A y1 —The apparent acceleration signal output by the normal acceleration gauge (rocket body coordinate system);

[0049] Φ Ay1 (z) — Attitude angle and overload feedback correction network;

[0050] K P K I —The ratio and integral coefficient of overload feedback.

[0051] (2) Design an extended state observer with adaptive bandwidth parameter variation

[0052] Because the control system faces environmental / intrinsic uncertainties and time-varying unknown disturbances, these factors can adversely affect its overload control accuracy, thus impacting the final load reduction effect. Therefore, further robust enhancement schemes to improve overload control performance are necessary to improve load reduction efficiency. Considering both robustness and practicality, an extended state estimation and compensation approach is adopted to improve load reduction control effectiveness. The basic extended state observer design is as follows:

[0053]

[0054] In the formula, y = ω z1 Indicates system output, ω z1 z1 and z2 represent the extended state observer's estimates of the system state and uncertain dynamics, respectively; b0 represents the control efficiency coefficient estimate; u represents the total control output of the controller; β1 > 0, β2 > 0 represent the observer coefficients, typically β1 = 2β n ,β2=β n 2 , where β n >0 indicates the observer's basic bandwidth parameter.

[0055] In the initial observation phase, linear state observers exhibit significant deviations between the actual and estimated system state variables. Due to the high gain of the observer, the initial disturbance estimation output shows a large peak, which becomes more pronounced as the observer gain increases. To address this issue, this invention employs a linear state observer design method with adaptively varying bandwidth parameters, specifically constructing an adaptively extended state observer. In this case, the basic bandwidth parameter β... n Adaptation is expressed as a nonlinear function in the following form:

[0056]

[0057] Where, β n0 >0 indicates the bandwidth limit value; κ>0 indicates the growth rate coefficient. The larger the value, the faster the curve rises, which means it is closer to the traditional design scheme.

[0058] Therefore, the traditional extended state observer will be improved into the following adaptive parametric form:

[0059]

[0060] With the introduction of an adaptive extended state observer, the original active load shedding control algorithm will be improved to the following form:

[0061]

[0062] (3) Design a parameter intelligent adaptive tuning strategy based on fuzzy inference

[0063] The improved active load shedding control algorithm consists of three parts: the first part is the attitude angle tracking control component, the second part is the overload tracking control component, and the first two parts balance the accuracy of programmed angle tracking and the load shedding control effect; the third part is the extended state observer compensation quantity, which aims to reduce the impact of system uncertainties on the load shedding control effect. Therefore, to address the "control capability competition" problem existing in the first two control components, this invention introduces an adaptive allocation coefficient. The system adaptively configures the capabilities of both components based on the current state, thereby achieving comprehensive performance optimization. Therefore, an active adaptive load reduction control algorithm is proposed:

[0064]

[0065] Regarding adaptive allocation coefficients To obtain the allocation coefficient, this invention will employ a fuzzy reasoning strategy to adaptively adjust the allocation coefficient online based on the current load reduction effect.

[0066] The concept of load margin is first introduced below, and its mathematical expression is defined as follows:

[0067]

[0068] Among them, Q α_max Q represents the aerodynamic load index value. α =q|α| represents the current aerodynamic load value, q represents the dynamic pressure, and α represents the angle of attack. When the calculated value is negative, Q is... α_YD Set it to 0. Therefore, when Q... α Stay away from Q α_max When the load margin is large, when Q α Approaching Q α_max When Q decreases, the load margin value decreases. α Greater than or equal to Q α_max At that time, the load margin becomes 0.

[0069] Because attitude angle deviation feedback and overload feedback share the launch vehicle's control capabilities, when the load margin Q α_YD If too small, increase the size appropriately. The value achieves a further load reduction effect when the load margin Q α_YD When the value is large, reduce it appropriately. This allows for the release of more control capabilities for attitude angle deviation control. Furthermore, considering that fuzzy control is an intelligent control method that does not require an accurate object model, incorporating human control experience and knowledge, and is also an easily implemented control method in engineering, this invention will employ fuzzy inference to achieve [the desired control]. Values ​​are adjusted online adaptively.

[0070] Load margin Q α_YD As the input signal to the fuzzy controller, the adaptive parameters will be... As the output signal of the fuzzy controller, the load margin is divided into four fuzzy regions: PB (positive large), PM (positive medium), PS (positive small), and ZO (zero value). It is divided into three fuzzy regions: PB, PM and PS.

[0071] Based on the following principle: when the load margin Q α_YD When the value is large, the allocation of capability to control attitude angle deviation can be prioritized, therefore, the value should be reduced. Value; when the load margin Q α_YD When the value is small, the allocation of overload feedback control capacity can be prioritized, therefore increasing it at this time... Value; when the load margin Q α_YD When the value is too small or even has no margin, it can be Set the value to the maximum level.

[0072] The final fuzzy mapping rules are as follows:

[0073] Table 1 Fuzzy Mapping Rules

[0074]

[0075] The table above contains four fuzzy IF-THEN mapping rules, with the rule format shown below:

[0076] R (j) :IF Q α_YD is A (j) THEN K IGC is B (j)

[0077] Among them, A (j) and B (j) Let Q represent the input and output fuzzy sets, respectively. The input and output fuzzy sets of the fuzzy controller are described using trigonometric functions, where the input variable Q... α_YD and output variables The domain ranges are [0.0, 1.0) and [Ukmin, Ukmax], respectively; Ukmin < 1.0 and Ukmax > 1.0 are the minimum and maximum values ​​of the domain of the output variable, respectively.

[0078] By employing the centroid method for defuzzification, the adaptive coefficients can be obtained through the following fuzzy logic reasoning mechanism:

[0079]

[0080] The active adaptive load reduction control method for launch vehicles based on fuzzy reasoning proposed in this application can be applied not only to launch vehicles, but also to other aircraft that require load reduction in high wind areas, thereby reducing the impact of aerodynamic loads on the structure of the launch vehicle during flight in high wind areas.

[0081] Example 2

[0082] This application discloses a mathematical simulation verification and comparative experiment to conduct mathematical simulation verification of the control scheme of the present invention, and further concludes the advantages of the load reduction effect of the present invention by comparing multiple control schemes.

[0083] Under typical deviation combinations, the aerodynamic load variation curves of the basic load reduction control scheme, the load reduction control scheme with an extended state observer, and the adaptive load reduction control scheme proposed in this invention are compared as follows: Figure 1 As shown.

[0084] In summary, during the high-wind flight segment (30-40s), the load reduction effect was significantly improved compared to the basic control scheme after adding the adaptive extended state observer. Furthermore, the load reduction effect was further improved to some extent after adding the intelligent parameter tuning controller. Specifically, the maximum aerodynamic load values ​​under the three load reduction control schemes were 228185.831, 188831.839, and 144136.100, respectively, demonstrating the load reduction effects of the adaptive extended state observer and the intelligent parameter tuning controller, thus verifying the effectiveness of the control method of this invention.

[0085] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A launch vehicle active adaptive offload control method based on fuzzy inference, characterized in that, Includes the following steps: (1) Design an active load reduction control scheme for the launch vehicle, including a combination of rocket attitude feedback and overload feedback. The traditional load reduction control equation is as follows: ； wherein represents the pitch attitude angle deviation under the arrow system; represents the attitude angle feedback gain coefficient; represents the visual acceleration signal output by the normal acceleration table; , respectively represent the attitude angle and overload feedback correction networks; , respectively represent the proportional and integral coefficients of the overload feedback; (2) Design an extended state observer with adaptive bandwidth parameter, including extended state estimation and compensation methods; ① The basic extended state observer is as follows: ； In the formula, Indicates system output, Indicates pitch angular velocity; and These represent the estimators of the system state and the uncertain dynamics by the extended state observer, respectively. This represents an estimate of the control efficiency coefficient; This indicates the overall control output of the controller; Represents the observer coefficients, typically taken as... ,in This represents the basic bandwidth parameter of the observer; ② Constructing adaptive extended state observer, using the linear state observer design method with adaptive bandwidth parameter change, the basic bandwidth parameter Adaptive, expressed as the following nonlinear function form: ； wherein, represents a bandwidth clipping value; represents a growth rate coefficient, the greater the value, the faster the curve rises, i.e. the closer to the conventional design. ③ The traditional extended state observer will be improved into the following adaptive parametric form: ； With the introduction of an adaptive extended state observer, the original active load shedding control algorithm will be improved to the following form: ； (3) Design a parameter intelligent adaptive tuning strategy based on fuzzy inference: A. After determining the expansion state observer compensation quantity, an adaptive distribution coefficient is introduced The active adaptive load shedding control algorithm is as follows: ； B. Load margin As the input signal of the fuzzy controller, the adaptive parameter As the output signal of the fuzzy controller, the load margin is divided into four fuzzy regions, respectively PB, PM, PS and ZO, which represent positive big, positive medium, positive small and zero value, and the parameter is divided into three fuzzy regions, respectively PB, PM and PS; C. Design fuzzy mapping rules, including 4 fuzzy IF-THEN mapping rules, with the following rule format: : IF is THEN is where, and denote the input and output fuzzy sets, respectively; the input and output fuzzy sets of the fuzzy controller are described by a triangular function, where the domain of the input variable and the domain of the output variable are [0.0, 1.0) and [Ukmin, Ukmax], respectively; Ukmin<1.0 and Ukmax>1.0 are the minimum and maximum values of the domain of the output variable, respectively; D. The centroid method is used for defuzzification. The adaptive coefficients are obtained through a fuzzy logic reasoning mechanism, and the calculation method is as follows: 。 2. The launch vehicle active adaptive offload control method of claim 1, wherein, The load margin expression in step (3) is defined as follows: ； in, This indicates the aerodynamic load index value. This indicates the current aerodynamic load value. Indicates dynamic pressure. Indicates the angle of attack; when the calculated value is negative, it will be... Set to 0; therefore, when keep away At that time, the load margin value is relatively large, when near When the load margin value decreases, Greater than or equal to At that time, the load margin becomes 0.

3. The launch vehicle active adaptive offload control method of claim 1, wherein, The division method of step (3) includes: using fuzzy inference to divide the fuzzy region The value is adjusted online and adaptively. When the load margin is small, the value is appropriately increased to achieve the effect of further reducing the load. When the load margin is large, the value is appropriately reduced to release more control ability for attitude angle deviation control.

Images