Aircraft attitude control method, system, equipment and medium

Abstract

The invention relates to an aircraft attitude control method, system and device and a medium, the aircraft attitude control method is used for attitude control of a wide-area unpowered aircraft, and comprises the following steps: S101, using an MPC as an attitude controller; s102, taking the simplified state space as a state space model required by the attitude controller; S103, acquiring a state space deviation observed by an extended state observer and caused by simplifying the state space; s104, the observed state space deviation is superposed to the state space to serve as the input of the attitude controller; s105, using the state space deviation to construct a state space predicted by a preset number of times in the future through iterative optimization; and S106, constructing a loss function, solving an input enabling the loss function to be minimum by using a quadratic programming algorithm, and taking the input as an optimal rudder output. According to the method, different working conditions do not need to be independently designed, the design workload is reduced, the calculation amount is reduced, and real-time operation can be achieved in a flight control computer.

Description

Technical Field

[0001] The present invention relates to the technical field of attitude control of wide - domain unpowered aircraft. In particular, it relates to an aircraft attitude control method, system, device and medium. Background Art

[0002] Wide - domain aircraft have a large flight airspace and speed - domain envelope. The traditional method is to design corresponding controllers at different speeds and altitudes and select different controllers and control parameters according to the flight state. This method requires a large amount of work in the design stage to design controllers for different working conditions for selection during flight. When the aircraft is actually flying, affected by various external environments, if the actual working condition is quite different from the designed working condition and the designed controller cannot meet the actual needs, serious flight accidents will occur. How to design an attitude controller that cannot adapt to different working conditions and the computational complexity of the state - space algorithm is difficult to run in real - time is an urgent problem to be solved. Summary of the Invention

[0003] The present invention provides an aircraft attitude control method, system, device and medium to solve the problems that the attitude controller cannot adapt to different working conditions and the computational complexity of the state - space algorithm is difficult to run in real - time.

[0004] To achieve the above object, in the first aspect, the present invention relates to an aircraft attitude control method for attitude control of wide - domain unpowered aircraft, including:

[0005] S101 Using MPC as the attitude controller;

[0006] S102 Using a simplified state - space as the required state - space model of the attitude controller, taking the pitch angle θ and pitch - rate q two - dimensional signals accurately measured by the sensor as the state variables of the attitude controller, and the simplified state - space at least omits the aircraft speed, angle of attack and altitude;

[0007] S103 Obtaining the state - space deviation caused by simplifying the state - space observed by the extended state observer ESO;

[0008] S104 Superimposing the observed state - space deviation on the state - space as the input of the attitude controller;

[0009] S105 Using the state - space deviation to construct the state - space predicted for a preset number of future times through iterative optimization;

[0010] S106 Constructing a loss function and using the quadratic programming algorithm to find the input that minimizes the loss function as the optimal rudder output.

[0011] Preferably, it includes: obtaining the state space deviation caused by simplifying the state space observed by the Extended State Observer (ESO), including:

[0012] Obtaining the linearization formula of the pitch angle θ and the pitch rate q: Where

[0013] M q Is the partial derivative of the pitch moment M with respect to the pitch rate q, and M δe Is the partial derivative of the pitch moment M with respect to the elevator deflection δ e Of;

[0014] The state space is:

[0015] The influence brought by the velocity V, the angle of attack α, the altitude H, and other state variables that affect the angles and angular rates and are omitted is collectively referred to as the deviation f, and the linearization formula can be described as:

[0016]

[0017] Superimposing the observed state space deviation onto the state space as the input of the attitude controller, specifically:

[0018] Taking the elevator deflection δ e As the system input u, adding a third state x 3 = f, to obtain the state space: Where, the first state x 1 Is the pitch angle θ, the second state x 2 Is the angular rate q, and the derivative of the deviation The output y = θ = x 1 , and f is the state space deviation;

[0019] It also includes introducing an error feedback coefficient L to ensure the stability of the Extended State Observer (ESO) and to estimate the state space deviation in real time, specifically:

[0020] Constructing the Extended State Observer (ESO) as:

[0021] L is the feedback vector,

[0022] Configuring the observer poles into the same roots to obtain the feedback vector L:

[0023] s 3 +β 1 s 2 +β 2 s + β 3 =(s + w 0 )3 , β 1 = 3w 0 , where w 0 is

[0024] the width of the observer band, β 1 , β 2 , β 3 are the coefficients in front of the observer respectively, and s is a complex variable.

[0025] Preferably, constructing the state quantity of the future preset number of predictions through iterative optimization using the state space deviation includes:

[0026] Taking the state space with the state space deviation as a reference, predicting the future state n times. The state space with the state space deviation is:

[0027] Let the initial state of the prediction be x(k);

[0028] x(k + n|k) is the state of the subsequent nth prediction based on the kth moment, and X k = Mx(k) + FU k + P,

[0029] where:

[0030]

[0031] X k and M are both 2(n + 1)×1 dimensional vectors, F is a 2(n + 1)×n dimensional vector, U k is an n×1 dimensional vector, and P is a 2(n + 1)×1 dimensional vector.

[0032] Preferably, constructing the loss function and using the quadratic programming algorithm to find the input that minimizes the loss function as the optimal rudder output includes:

[0033] Defining R as the input command of the system Then the 2(n + 1)×1 dimensional input matrix R is composed of R k = (R R R...R) T ;

[0034] Defining Q as the weight matrix of the state quantities (θ, q) where a 1 is the weight coefficient of the state quantity θ, a 2 is the weight coefficient of the state quantity q, and the 2(n + 1)×2(n + 1) dimensional diagonal weight matrix can be composed of Q

[0035] Define I as the weight coefficient for the input, and an n×n weight diagonal matrix is obtained from the weight coefficient I

[0036] Construct the loss function: where E = Mx(k) - R k ,

[0037] Use the quadratic programming algorithm for the loss function to find the input that minimizes the loss function, and use it as the optimal rudder output for the current state of the aircraft:

[0038]

[0039] Preferably, it further includes S107: Use S103 to S106 in real time to obtain the optimal rudder output of the aircraft flying in the air at any current moment.

[0040] To achieve the above object, in a second aspect, the present invention relates to an aircraft attitude control system for attitude control of a wide - domain unpowered aircraft, including:

[0041] An attitude control module for using MPC as an attitude controller;

[0042] A simplified state - space module for using the simplified state - space as the required state - space model of the attitude controller, and taking the pitch angle θ and pitch - rate q two - dimensional signals accurately measured by the sensor as the state variables of the attitude controller. The simplified state - space omits at least the speed, angle of attack, and altitude of the aircraft;

[0043] An error acquisition module for acquiring the state - space deviation caused by simplifying the state - space observed by the extended state observer ESO;

[0044] An error superposition module for superposing the observed state - space deviation onto the state - space as the input of the attitude controller;

[0045] A prediction module for using the state - space deviation to construct the state variables predicted for a preset number of times in the future through iterative optimization;

[0046] An optimal rudder output module for constructing a loss function and using the quadratic programming algorithm to find the input that minimizes the loss function as the optimal rudder output.

[0047] To achieve the above object, in a third aspect, the present invention further relates to an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the computer program, the above-mentioned aircraft attitude control method is implemented.

[0048] To achieve the above object, in a fourth aspect, the present invention further relates to a computer-readable storage medium, in which instructions are stored, and when the instructions run, the above-mentioned aircraft attitude control method is executed.

[0049] An aircraft attitude control method, system, device and medium according to the present invention has the following beneficial effects compared with the prior art:

[0050] The present invention uses MPC (Model Predictive Control) as the attitude controller. Due to the limited resources of the flight control computer, in order to reduce the computational complexity, a simplified state space is used as the required state space model of the MPC, and the two-dimensional signals of the angle θ and the angular rate q that can be accurately measured by the sensor are used as state variables. In order to compensate for the error caused by the state simplification, an Extended State Observer (ESO) is used to observe the state space deviation caused by the omitted state, and the observed error amount is superimposed on the state space as the model input of the MPC.

[0051] The attitude controller designed by the present invention only needs to know the simplified state space of the aircraft, and obtains the optimal rudder output through iterative optimization, without the need for separate design for different working conditions, greatly reducing the design workload; an Extended State Observer is used to make up for the state space output deviation caused by model simplification and the external environment, making the linearized model of the aircraft more consistent with the actual situation, which is conducive to the MPC calculation to better adapt to the actual flight output rudder. The simplified state space can greatly reduce the computational complexity of the algorithm, and it can be implemented in real time in the flight control computer. The controller inputs are two command signals of the angle θ and the angular rate q, and the angular rate command uses proportional control to further correct the control error and reduce the static error. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is the method flow of an aircraft attitude control method in Embodiment 1 of the present invention Figure 1 ;

[0053] Figure 2 is the method flow of an aircraft attitude control method in Embodiment 1 of the present invention Figure 2 ;

[0054] Figure 3 is the attitude angle response curve of an aircraft attitude control method in Embodiment 1 of the present invention;

[0055] Figure 4It is the speed change curve of an aircraft attitude control method in Embodiment 1 of the present invention;

[0056] Figure 5 It is the altitude change curve of an aircraft attitude control method in Embodiment 1 of the present invention;

[0057] Figure 6 It is the structural schematic diagram of an aircraft attitude control system in Embodiment 2 of the present invention;

[0058] Figure 7 It is the structural schematic diagram of an electronic device in Embodiment 3 of the present invention. Detailed implementation manners

[0059] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. In addition, it should be noted that, for the sake of convenience of description, only parts related to the present invention rather than all structures are shown in the drawings.

[0060] Embodiment 1

[0061] An aircraft attitude control method, please refer to Figures 1-5 , which is used for the attitude control of a wide - area unpowered aircraft and is implemented by a flight control computer. As Figure 1 shown, it includes the following steps: S101 to S106.

[0062] S101 uses MPC (Model Predictive Control) as the attitude controller.

[0063] The simplified state space is used as the required state - space model of the attitude controller. The two - dimensional signals of the pitch angle θ and the pitch - rate q accurately measured by the sensor are used as the state variables of the attitude controller. The simplified state space omits at least the speed, angle of attack, and altitude of the aircraft, and some other state variables that have an impact on the angle and angular rate can also be omitted.

[0064] Due to the limited resources of the flight control computer, in order to reduce the computational amount, the simplified state space is used as the required state - space model of the MPC.

[0065] The inputs of the attitude controller are two command signals of the angle θ and the angular rate q. The angular - rate command uses proportional control to further correct the control error and reduce the static error.

[0066] S103 obtains the state - space deviation caused by the simplified state space observed by the Extended State Observer (ESO).

[0067] In this embodiment, in order to compensate for the error caused by state simplification, an Extended State Observer (ESO) is used to observe the state - space deviation caused by the omitted states, and the observed error amount is superimposed on the state space as the model input of the MPC.

[0068] S103 specifically includes S1031 - S1032:

[0069] S1031 obtains the linearized formula of the pitch angle θ and the pitch - angle rate q: where M q is the partial derivative of the pitch moment M with respect to the pitch - angle rate q, that is: M δe is the partial derivative of the pitch moment M with respect to the elevator amount δ e , that is:

[0070] The state space is: This state space does not include the velocity V, the angle of attack α, the altitude H, and other state variables that affect the angle and the angular rate.

[0071] S1032 collectively refers to the influence brought by the velocity V, the angle of attack α, the altitude H, and other omitted state variables that affect the angle and the angular rate as the state - space deviation f. The linearized formula can be described as:

[0072]

[0073] The state - space deviation f is a function of the velocity V, the angle of attack α, the altitude H, and other omitted state variables that affect the angle and the angular rate, f = f(V, α, H,...).

[0074] S104 superimposes the observed state - space deviation on the state space as the input of the attitude controller.

[0075] Specifically, taking the elevator amount δ e as the system input u, adding the third state x 3 = f, the state space obtained is:

[0076] where the first state x 1 is the pitch angle θ, the second state x 2 is the angular rate q, the derivative of the deviation the output y = θ = x 1 , and f is the state - space deviation.

[0077] It also includes introducing an error feedback coefficient L to ensure the stability of the Extended State Observer (ESO) and to perform real - time estimation of the state - space deviation. Specifically:

[0078] The state space can be simplified to:

[0079] Among them:

[0080]

[0081] Construct the extended state observer ESO as:

[0082] L is the feedback vector;

[0083] Configure the observer poles into the same roots to obtain the feedback vector L:

[0084] s 3 +β 1 s 2 +β 2 s+β 3 =(s + w 0 ) 3 , β 1 = 3w 0 , where w 0 is

[0085] the observer bandwidth, β 1 、β 2 、β 3 are the coefficients before the observer respectively, and s is a complex variable.

[0086] The specific derivation process is as follows, and thus the extended differential equation can be written as:

[0087]

[0088] Convert it into the state - space form:

[0089]

[0090] y = Cx

[0091] where:

[0092]

[0093] The following - form state observer can be constructed:

[0094]

[0095] By introducing the error - feedback coefficient, both the stability of the observer can be guaranteed and the purpose of real - time estimation of the deviation in the system can be achieved. The feedback vector L is obtained by the method of configuring the observer poles into the same roots, that is:

[0096] s 3 +β1 s 2 +β 2 s + β 3 =(s + w 0 ) 3

[0097] β 1 =3w 0 ,

[0098] The feedback vector L can be expressed as:

[0099] where w 0 is the observer bandwidth, and β 1 , β 2 , β 3 are the coefficients before the observer respectively, and s is a complex variable.

[0100] S105 constructs the state quantity for future preset number of predictions through iterative optimization using the state - space deviation.

[0101] Specifically, it includes:

[0102] Taking the state - space with state - space deviation as the basis, predict the future state n times. The state - space with state - space deviation is:

[0103] Let the initial state of the prediction be x(k);

[0104] x(k + n|k) is the state of the subsequent nth prediction based on the k - th moment, and we get X k =Mx(k)+FU k +P,

[0105] where:

[0106]

[0107] X k and M are both 2(n + 1)×1 - dimensional vectors, F is a 2(n + 1)×n - dimensional vector, U k is an n×1 - dimensional vector, and P is a 2(n + 1)×1 - dimensional vector.

[0108] In this embodiment, the specific prediction derivation process is as follows:

[0109] Using the state observer constructed in the first section, the deviation term f can be obtained. Introduce f into the state - space for model prediction.

[0110]

[0111] where,

[0112] Then the above state space can be written as:

[0113]

[0114] Using this state space as a reference, make n predictions about future states. Let the initial state of the prediction be x(k); x(k + n|k) is the nth subsequent prediction based on the kth moment.

[0115] Current state: x(k|k) = x(k)

[0116] First prediction:

[0117]

[0118] Second prediction:

[0119]

[0120] Third prediction:

[0121]

[0122] nth prediction:

[0123]

[0124] Write it in the form of a state space:

[0125]

[0126] It can be abbreviated as:

[0127] X k = Mx(k) + FU k + P

[0128] Where:

[0129]

[0130] X k And M are both 2(n + 1)×1 dimensional vectors; F is a 2(n + 1)×n dimensional vector; U k is an n×1 dimensional vector; P is a 2(n + 1)×1 dimensional vector.

[0131] S106 Construct a loss function, and use the quadratic programming algorithm to find the input that minimizes the loss function as the optimal rudder output.

[0132] Among them, constructing the loss function includes:

[0133] Define R as the input command of the system Then, an input matrix \(R\) of dimension \(2(n + 1)\times1\) can be formed by \(R\). k =(RR R...R) T ;

[0134] Define \(Q\) as the weight matrix of the state variables \((\theta,q)\). where \(a\) 1 is the weight coefficient of the state variable \(\theta\), and \(a\) 2 is the weight coefficient of the state variable \(q\). A diagonal weight matrix of dimension \(2(n + 1)\times2(n + 1)\) can be formed by \(Q\).

[0135] Define \(I\) as the weight coefficient of the input. Then, an \(n\times n\) weight diagonal matrix can be obtained from the weight coefficient \(I\).

[0136] Construct the loss function:

[0137] where \(E = Mx(k)-R\). k ,

[0138] Use the quadratic programming algorithm to find the input that minimizes the loss function as the optimal rudder output, including:

[0139] Use the quadratic programming algorithm for the simplified loss function to find the input \(U_{kAct}\) that minimizes \(J\) as

[0140] Then, a set of elevator amounts optimal for the current state of the aircraft is obtained: \(\delta\). e =U kAct (1)=u(k).

[0141] In this embodiment, the first value \(u(k)\) in \(U\) kAct is taken as the current output value and given to the aircraft elevator amount.

[0142] In this embodiment, the specific final loss function can be simplified to the standard form The derivation process is as follows:

[0143] From the \(n\times n\) weight diagonal matrix defined above Construct the loss function \(J(U_{K})\).

[0144]

[0145] Let \(E = Mx(k)-R\). k , then

[0146] Since is a constant and does not affect the final result, so it can be ignored. Let Then the final loss function can be simplified to the standard form

[0147]

[0148] In some embodiments, after S106, S107 (not shown in the drawings) is further included: using S103 to S106 in real time to obtain the optimal rudder output of the aircraft flying in the air at any current moment.

[0149] To better illustrate the invention, as Figures 3-5 shown below is an example:

[0150] The following is a certain type of unpowered aircraft. Using the ESO-MPC shown in Example 1 of the present invention, a step command for the attitude of 5 degrees of pitch angle is given to it, and its step response is viewed.

[0151] In the case of unpowered, during the attitude step response flight for 20 s, the speed drops by 25 m / s, the altitude drops by 800 m, the pitch attitude step response has no overshoot and can achieve stable tracking, and the effect is excellent, achieving good results on a wide-range unpowered aircraft.

[0152] Embodiment 2

[0153] An aircraft attitude control system for attitude control of a wide-range unpowered aircraft. In this embodiment, a flight control computer is used. Please refer to Figure 6 , including an attitude control module 61, a simplified state space module 62, an error acquisition module 63, an error superposition module 64, a prediction module 65, and an optimal rudder output module 66.

[0154] The attitude control module 61 is used to use MPC as an attitude controller;

[0155] The simplified state space module 62 is used to use the simplified state space as the required state space model of the attitude controller, and uses the two-dimensional signals of the pitch angle θ and the pitch rate q accurately measured by the sensor as the state quantities of the attitude controller. The simplified state space omits at least the speed, angle of attack, and altitude of the aircraft;

[0156] The error acquisition module 63 is used to acquire the state space deviation caused by the simplified state space observed by the extended state observer ESO;

[0157] The error superposition module 64 is used to superpose the observed state space deviation onto the state space as the input of the attitude controller;

[0158] The prediction module 65 is used to construct the state quantities predicted for a preset number of times in the future through iterative optimization using the state space deviation;

[0159] The optimal rudder output module 66 is used to construct a loss function and use the quadratic programming algorithm to find the input that minimizes the loss function as the optimal rudder output.

[0160] In some embodiments, the error acquisition module 63 is specifically configured to:

[0161] Obtain the linearization formula of the pitch angle θ and the pitch rate q: Where

[0162] M q is the partial derivative of the pitch moment M with respect to the pitch rate q, and M δe is the partial derivative of the pitch moment M with respect to the elevator deflection δ e ;

[0163] The state space is:

[0164] Collectively refer to the influences brought by the speed V, the angle of attack α, the altitude H, and other state variables that are omitted and affect the angles and angular rates as the deviation f. The linearization formula can be described as:

[0165]

[0166] The error superposition module 64 is specifically configured to: Use the elevator deflection δ e as the system input u, add the third state x 3 = f, and obtain the state space as: Where the first state x 1 is the pitch angle θ, the second state x 2 is the angular rate q, and the derivative of the deviation The output y = θ = x 1 , and f is the state space deviation;

[0167] It also includes: an error feedback coefficient introduction module 67 (not shown in the figure) connected to the error acquisition module 63, which is used to introduce an error feedback coefficient L to ensure the stability of the extended state observer ESO;

[0168] Introducing the error feedback coefficient L is specifically:

[0169] Construct an extended state observer ESO:

[0170] L is the feedback vector,

[0171] Configure the observer poles into the same roots to obtain the feedback vector L:

[0172] s 3 +β 1 s 2 +β2 s + β 3 = (s + w 0 ) 3 , β 1 = 3w 0 ,

[0173] where w 0 is the observer bandwidth, and β 1 , β 2 , β 3 are the coefficients before the observer respectively, and s is a complex variable.

[0174] In some embodiments, the prediction module 65 is specifically configured to:

[0175] Use the state space with state space deviation as a reference to make n predictions on the future state. The state space with state space deviation is:

[0176] Let the initial state of the prediction be x(k);

[0177] x(k + n|k) is the state of the subsequent nth prediction based on the kth moment, and X k = Mx(k) + FU k + P,

[0178] where:

[0179]

[0180] X k and M are both 2(n + 1)×1 dimensional vectors, F is a 2(n + 1)×n dimensional vector, U k is an n×1 dimensional vector, and P is a 2(n + 1)×1 dimensional vector.

[0181] In some embodiments, the optimal rudder output module 66 is specifically configured to:

[0182] Define R as the input command of the system Then form a 2(n + 1)×1 dimensional input matrix R from R k = (R R R...R) T ;

[0183] Define Q as the weight matrix of the state variables (θ, q) where a 1 is the weight coefficient of the state variable θ, and a 2 is the weight coefficient of the state variable q. The 2(n + 1)×2(n + 1) dimensional diagonal weight matrix can be formed by Q

[0184] Define I as the weight coefficient of the input, and an n×n weight diagonal matrix is obtained from the weight coefficient I

[0185] Construct the loss function: where E = Mx(k) - R k ,

[0186] Use the quadratic programming algorithm for the loss function to find the input that minimizes J Obtain a set of optimal elevator amounts for the current state of the aircraft, and take U kAct Take the first value u(k) in e = U kAct (1) = u(k) as the current output value to the aircraft elevator δ

[0187] It also includes a real-time optimization module 68 (not shown in the figure), which is used to execute the error acquisition module 63, error superposition module 64, prediction module, and optimal rudder amount output module 65 in real time, and obtain the optimal rudder amount output for the aircraft flying in the air at any current moment

[0188] A flight vehicle attitude control system according to this embodiment has the same implementation process, method, and effect as the flight vehicle attitude control method described in Embodiment 1, and will not be elaborated here

[0189] Embodiment 3

[0190] As Figure 7 shown, this embodiment relates to an electronic device including at least one processor and a memory communicatively connected to at least one processor. Among them, the memory stores a computer program that can be run by at least one processor, and the computer program is executed by at least one processor so that at least one processor can execute a flight vehicle attitude control method of Embodiment 1 and achieve the corresponding beneficial effects of a flight vehicle attitude control method, which will not be elaborated here. The electronic device provided in this embodiment can be a personal computer, such as a desktop computer, all-in-one computer, notebook computer, tablet computer, etc., and can also be a terminal device such as a mobile phone, wearable device, handheld computer, etc. In this embodiment, the electronic device is a flight control computer. The electronic device is only an example and should not bring any limitations to the functions and usage scope of the embodiments of the present invention

[0191] The components of the electronic device 3 may include, but are not limited to: the above-mentioned at least one processor 4, the above-mentioned at least one memory 5, and a bus 6 connecting different system components (including the memory 5 and the processor 4)

[0192] The bus 6 includes a data bus, an address bus, and a control bus

[0193] The memory 5 may include volatile memory, such as random access memory (RAM) 51 and / or cache memory 52, and may further include read-only memory (ROM) 53.

[0194] The memory 5 may also include a program / utilities 55 having a set (at least one) of program modules 54. Such program modules 54 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment.

[0195] The processor 4 executes various functional applications and data processing by running computer programs stored in the memory 5, such as the above-mentioned aircraft attitude control method.

[0196] The electronic device 3 may also communicate with one or more external devices 7 (such as a keyboard, pointing device, etc.). Such communication may be performed through an input / output (I / O) interface 8. Also, the electronic device 3 may communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 9. As Figure 5 shown, the network adapter 9 communicates with other modules of the electronic device 3 through the bus 6. It should be understood that although Figure 5 not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 3, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (redundant array of independent disks) systems, tape drives, and data backup storage systems, etc.

[0197] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present invention, the features and functions of two or more of the above-described units / modules may be embodied in one unit / modules. Conversely, the features and functions of one unit / modules described above may be further divided and embodied by multiple unit / modules.

[0198] Embodiment 4

[0199] The present invention relates to a computer-readable storage medium storing instructions that, when executed, implement an aircraft attitude control method according to Embodiment 1. The implementation process, method, and effects are the same as those of the aircraft attitude control method described in Embodiment 1 and will not be repeated here.

[0200] It should be noted that in this text, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements not only includes those elements but also other elements not explicitly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the presence of additional identical elements in the process, method, article or device comprising such element.

[0201] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A method for controlling an aircraft attitude, characterized in that: For attitude control of wide-range unpowered aircraft, including: S101 uses MPC as attitude controller; S102: a simplified state space is used as the required state space model of the attitude controller, and the two-dimensional signals of the pitch angle θ and the pitch angle rate q accurately measured by the sensor are used as the state quantity of the attitude controller, wherein the simplified state space at least omits the speed, the angle of attack and the altitude of the aircraft; S103: obtaining a state space deviation observed by an extended state observer (ESO) caused by simplifying the state space; S104 superimposes the observed state space deviation to the state space as an input of the attitude controller; S105 uses the state space deviation to construct a state space predicted for a preset number of future times through iterative optimization; S106 constructs a loss function, and uses a quadratic programming algorithm to find an input that minimizes the loss function as the optimal steering output.

2. The method for controlling the attitude of an aircraft according to claim 1, characterized in that: include: The step of obtaining the state space deviation observed by the extended state observer (ESO) caused by simplifying the state space comprises: The linearized formula of pitch angle θ and pitch angle rate q is obtained: Among them, M q is the partial derivative of the pitch moment M with respect to the pitch angular rate q, M δe is the pitch moment M on the elevator amount δ e The partial derivative of The state space is: The influence of velocity V, angle of attack α, height H and other omitted state variables that affect angle and angular rate are collectively referred to as deviation f. The linearization formula can be described as: The step of superimposing the observed state space deviation to the state space as the input of the attitude controller is specifically as follows: The elevator amount δ e As the system input u, add the third state x3=f, and the state space is: Among them, the first state x1 is the pitch angle θ, the second state x2 is the angular rate q, and the derivative of the deviation Output y = θ = x1, f is the state space deviation; It also includes introducing an error feedback coefficient L to ensure the stability of the extended state observer ESO and to estimate the state space deviation in real time, specifically: The extended state observer ESO is constructed as: L is the feedback vector, Arrange the observer poles to the same root to obtain the feedback vector L: s 3 +β1s 2 +β2s+β3=(s+w0) 3 ,β1=3w0, in w0 is the observer bandwidth, β1, β2, β3 are the coefficients before the observer, and s is a complex variable.

3. The method for controlling the attitude of an aircraft according to claim 2, characterized in that: The method of using the state space deviation to construct a state quantity predicted for a preset number of times in the future through iterative optimization includes: Using the state space with state space deviation as a reference, n predictions are made for the future state. The state space with state space deviation is: Let the initial state of prediction be x(k); x(k+n|k) is the state of the subsequent n-th prediction based on time k, and we get X k =Mx(k)+FU k +P, in: X k and M are all 2(n+1)×1 dimensional vectors, F is a 2(n+1)×n dimensional vector, U k is an n×1 dimensional vector, and P is a 2(n+1)×1 dimensional vector.

4. The method for controlling the attitude of an aircraft according to claim 3, characterized in that: The loss function is constructed, and a quadratic programming algorithm is used to find an input that minimizes the loss function as the optimal steering quantity output, including: Define R as the input command of the system Then R is composed of a 2(n+1)×1 dimensional input matrix R k =(RR R...R) T ; Define Q as the weight matrix of the state (θ, q) Where a1 is the weight coefficient of the state quantity θ, a2 is the weight coefficient of the state quantity q, and Q can form a 2(n+1)×2(n+1) dimensional diagonal weight matrix Define I as the weight coefficient for the input, then the n×n weight diagonal matrix is ​​obtained from the weight coefficient I Construct the loss function: Where E = Mx(k)-R k , The quadratic programming algorithm is used to find the input that minimizes the loss function, which is used as the optimal steering output for the current state of the aircraft:

5. The method for controlling the attitude of an aircraft according to claim 1, characterized in that: The process also includes S107: using S103 to S106 in real time to obtain the optimal steering amount output for the aircraft flying in the air at any current time.

6. An aircraft attitude control system, characterized in that: For attitude control of wide-range unpowered aircraft, including: Attitude control module, used to use MPC as attitude controller; A simplified state space module is used for simplifying the state space as the state space model required by the attitude controller, and using the two-dimensional signals of the pitch angle θ and the pitch angle rate q accurately measured by the sensor as the state quantity of the attitude controller, wherein the simplified state space at least omits the speed, angle of attack and altitude of the aircraft; An error acquisition module, used for acquiring a state space deviation observed by an extended state observer (ESO) caused by simplifying the state space; an error superposition module, used for superimposing the observed state space deviation into the state space as an input of the attitude controller; A prediction module, used to construct a state quantity predicted for a preset number of future times by iterative optimization using the state space deviation; The optimal steering quantity output module is used to construct a loss function and use a quadratic programming algorithm to find the input that minimizes the loss function as the optimal steering quantity output.

7. An aircraft attitude control system according to claim 6, characterized in that: The error acquisition module is specifically used for: The linearized formula of pitch angle θ and pitch angle rate q is obtained: in M q is the partial derivative of the pitch moment M with respect to the pitch angular rate q, M δe is the pitch moment M on the elevator amount δ e The partial derivative of The state space is: The influence of velocity V, angle of attack α, height H and other omitted state variables that affect angle and angular rate are collectively referred to as deviation f. The linearization formula can be described as: The error superposition module is specifically used to: e As the system input u, add the third state x3=f, and the state space is: Among them, the first state x1 is the pitch angle θ, the second state x2 is the angular rate q, and the derivative of the deviation Output y = θ = x1, f is the state space deviation; It also includes: an error feedback coefficient introduction module, used to introduce an error feedback coefficient L to ensure the stability of the extended state observer ESO; The introduced error feedback coefficient L is specifically: Construct the extended state observer ESO: L is the feedback vector, Arrange the observer poles to the same root to obtain the feedback vector L: s 3 +β1s 2 +β2s+β3=(s+w0) 3 ,β1=3w0, among them w0 for The observer bandwidth, β1, β2, β3 are the coefficients before the observer, and s is a complex variable.

8. An aircraft attitude control system according to claim 7, characterized in that: The prediction module is specifically used for: Using the state space with state space deviation as a reference, n predictions are made for the future state. The state space with state space deviation is: Let the initial state of prediction be x(k); x(k+n|k) is the state of the subsequent n-th prediction based on time k, and we get X k =Mx(k)+FU k +P, in: X k and M are all 2(n+1)×1 dimensional vectors, F is a 2(n+1)×n dimensional vector, U k is an n×1 dimensional vector, and P is a 2(n+1)×1 dimensional vector.

9. An aircraft attitude control system according to claim 8, characterized in that: The optimal steering quantity output module is specifically used for: Define R as the input command of the system Then R is composed of a 2(n+1)×1 dimensional input matrix R k =(RR R...R) T ; Define Q as the weight matrix of the state (θ, q) Where a1 is the weight coefficient of the state quantity θ, a2 is the weight coefficient of the state quantity q, and Q can form a 2(n+1)×2(n+1) dimensional diagonal weight matrix Define I as the weight coefficient for the input, then the n×n weight diagonal matrix is ​​obtained from the weight coefficient I Construct the loss function: Where E = Mx(k)-R k , Use the quadratic programming algorithm to find the input that minimizes J for the loss function Get a set of optimal elevator inputs for the current state of the aircraft, and take U kAct The first value u(k) in is given as the current output value to the aircraft elevator δ e =U kAct (1) = u(k). It also includes a real-time optimization module for executing the error acquisition module, the error superposition module, the prediction module and the optimal steering amount output module in real time to obtain the optimal steering amount output of the aircraft flying in the air at any current time.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, an aircraft attitude control method according to any one of claims 1 to 5 is implemented.

11. A computer-readable storage medium, characterized in that: The storage medium stores instructions, which, when executed, execute an aircraft attitude control method as described in any one of claims 1-5.

Images