Preset index control method based on intelligent adaptive boundary adjustment

Abstract

The invention discloses a preset index control method based on intelligent adaptive boundary adjustment, and the method comprises the following steps: designing a basic form of a preset performance boundary function, and obtaining a control equation in a constrained state according to a kinetic equation; selecting a state quantity capable of influencing the control performance, and applying weight distribution to parameter values in a preset performance boundary function and subsequent control gains in combination with the full-feedback adaptive recurrent neural network model; carrying out equivalent unconstrained conversion on the constrained control model based on the obstacle function to obtain an unconstrained equivalent control model based on the auxiliary variable; a controller is designed according to a preset time stability theory and a backstepping method, so that the design of a preset index controller which adjusts and controls the response speed only through one parameter and can restrain the control performance is completed, and a control gain can be set in advance according to a flight task and a corresponding performance requirement thereof. The method can give full play to the ability of the aircraft, and improves the control quality of any flight stage under the full flight profile.

Description

Technical Field

[0001] The present invention relates to a method for controlling preset indicators of an aircraft, and more particularly to a method for controlling preset indicators based on intelligent adaptive boundary adjustment. Background Art

[0002] During the flight of a variant aircraft, due to its large flight span, diverse mission types, and different multi-mission objective requirements, the design of its controller often involves compromise. However, the compromise-designed controller cannot fully utilize the capabilities of the aircraft, reducing the advantages of variant flight. In addition, during flight, due to fixed control parameters, it is unable to address the problem of changing control performance requirements caused by sudden mission changes. Therefore, there is an urgent need for a control method that can intelligently and adaptively adjust the control performance boundary according to the actual flight state and can preset performance indicators. Summary of the Invention

[0003] Aiming at the problems that the traditional aircraft control system cannot accurately design control performance indicators in advance, cannot change control parameters according to real-time flight states and mission requirements, optimize control performance, and fully utilize the flight capabilities of the aircraft, the present invention provides a method for controlling preset indicators based on intelligent adaptive boundary adjustment. This method can not only preset control performance indicators in advance, but also adaptively adjust the control performance boundary and control gain according to the real-time flight state and mission requirements of the aircraft, so as to fully utilize the capabilities of the aircraft itself and improve the control quality at any flight stage under the full flight profile.

[0004] The objectives of the present invention are achieved through the following technical solutions:

[0005] A method for controlling preset indicators based on intelligent adaptive boundary adjustment includes the following steps:

[0006] Step 1: Design the basic form of the preset performance boundary function according to the mission scenario and corresponding performance requirements, and obtain the control equation under constrained states according to the dynamic equation;

[0007] Step 2: Select the state variables that can affect control performance among the real-time flight state, deformation state, and multi-mission requirements, and combine the full-feedback adaptive recurrent neural network model to assign weights to the parameter values in the preset performance boundary function and subsequent control gains, so as to achieve the purpose of adaptively adjusting control performance;

[0008] Step 3: Perform an equivalent unconstrained transformation on the constrained control model based on the barrier function to obtain an unconstrained equivalent control model based on auxiliary variables, thus facilitating the design of the controller;

[0009] Step 4: Design a controller based on the preset time stability theory and the backstepping method, so as to complete the design of a preset index controller that can adjust the control response speed with only one parameter and can constrain the preset indexes of the control performance, and the control gain can be set in advance according to the flight mission and its corresponding performance requirements.

[0010] Compared with the prior art, the present invention has the following advantages:

[0011] (1) By adopting the intelligent adaptive boundary adjustment technology, the preset performance boundary of the controller can be adjusted according to the real-time flight state and mission requirements of the aircraft, giving full play to the real-time flight ability and improving the control quality at any moment of the entire flight profile.

[0012] (2) Based on the double-limit weighting theory, a controller that can adjust the response speed with only one control parameter is designed. At the same time, combined with the preset performance function, the pre-defined design of the control system ability can be carried out quickly and conveniently, which has good engineering practicability. Description of the Drawings

[0013] Figure 1 It is a flowchart of a preset index control method based on intelligent adaptive boundary adjustment. Detailed Embodiment

[0014] The technical solutions of the present invention will be further described below in conjunction with the drawings, but are not limited thereto. Any modification or equivalent replacement of the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention shall be covered by the protection scope of the present invention.

[0015] The present invention provides a preset index control method based on intelligent adaptive boundary adjustment. As Figure 1 shown, taking a morphing aircraft as an example, the method includes the following steps:

[0016] Step 1: Design the basic form of the preset performance boundary function according to the mission scenario and the corresponding performance requirements, and obtain the control equation under the constrained state according to the dynamic equation. The specific steps are as follows:

[0017] Step 1.1: Analyze the control accuracy ρ ∞ , overshoot σ a and response time T a required by the control system according to the mission scenario, and design the basic form of the preset performance boundary function ρ:

[0018]

[0019] where ρ0 is the initial value of the performance function, ρ ∞ is the final value of the performance function, t is time, and τ is any small constant.

[0020] Step 1.2: According to the aircraft dynamics equation (2), integrate the state variables to obtain the aircraft state space equation (3) for the surface control system design:

[0021]

[0022]

[0023]

[0024] where, Ω = [σ, β, α] T represents the Euler angle vector, and σ, β, α are the bank angle, sideslip angle, and angle of attack respectively; ω = [ω x , ω y , ω z T represents the attitude angular velocity vector, and ω x , ω y , ω z are the roll angular velocity, yaw angular velocity, and pitch angular velocity respectively; ρ1 = [ρ 11 , ρ 12 , ρ 13 represents the constraint on the system state variable x1; ρ2 = [ρ 21 , ρ 22 , ρ 23 represents the constraint on the system state variable x2; J represents the inertia matrix; R represents the attitude transformation matrix; d is the unknown disturbance; M is the torque vector; the expressions of R, J, ω × are:

[0025] [[ID=!47]]

[0026] where, x1 = Ω, x2 = ω; M δc represents the control command vector, M Δδ represents the torque caused by the aerodynamic rudder deviation, and M0 represents the torque generated by the projectile body; f = -J -1 ω[[ID=!55]] × Jω + J -1 M0, D = J -1 d + J -1 M Λδ represents the total disturbance;

[0027] Thus, the state error equation is obtained as:

[0028]

[0029] where, e1 = x1 - x d , e2 = x2 - x 2d , x d represents the desired command of x1, x 2d It should be noted that there may be some inaccuracies in the translation due to the complexity and potential ambiguity of the original text, especially in the parts where the specific mathematical and technical meanings need to be precisely understood in the context of aerospace engineering. If possible, it is recommended to consult an expert in the relevant field for a more accurate interpretation.Instruction for representing the expectation of x2.

[0030] Step 2: Select the state variables that can affect the control performance among the real-time flight state, deformation state, and multi-task requirements, and combine the full-feedback adaptive recurrent neural network model to apply weight distribution to the parameter values in the preset performance boundary function and the subsequent control gains to achieve the purpose of adaptively adjusting the control performance. The specific steps are as follows:

[0031] Step 2.1: Select the real-time flight state Ω = [σ, β, α] T , the sweep angle Λ, the wingspan L, and the response time requirement and control accuracy requirement variables in the deformation state as the state variables for neural network training;

[0032] Step 2.2: Construct a neural network using radial basis functions and adjust the weight vector to obtain the value of the adaptive boundary state;

[0033]

[0034] Among them, is the weight coefficient matrix, φ(x) represents the activation function, and ρ a (x) is the adaptive adjustment boundary term, thus obtaining the adaptive boundary adjustment function:

[0035]

[0036] Step 3: Based on the barrier function, perform an equivalent unconstrained transformation on the constrained control model to obtain an unconstrained equivalent control model based on the auxiliary variable, thereby facilitating the design of the controller. The specific steps are as follows:

[0037] Select the barrier function in the form of:

[0038]

[0039]

[0040] Among them, ξ is the auxiliary error variable. Combining with Equation (6), the state error equation under unconstrained conditions can be obtained as:

[0041]

[0042] Step 4: Preset the time stability theory and backstepping method to design the controller, thereby completing the design of the preset index controller that can adjust the control response speed with only one parameter and can constrain the control performance, and the control gain can be set in advance according to the flight mission and its corresponding performance requirements. The specific steps are as follows:

[0043] Step 4.1: Design the virtual control quantity in the following form according to the first equation of Equation (11) and the preset time stability theory:

[0044]

[0045] Among them, \(0 < \kappa < 1\), \(T(x)\) is a convergence time adjustment term, which can be adaptively adjusted according to the flight state, and \(x\) 2d0 is an equivalent virtual control term, whose function is to eliminate the disturbance term outside the reaching law;

[0046] Step 4.2: Design the control quantity in the following form according to the second equation of Equation (11):

[0047]

[0048] Among them, \(M\) δ0 is an equivalent control term, whose function is to eliminate the disturbance term outside the reaching law, thereby completing the design of the controller.

[0049] According to the control flow Figure 1 It can be seen that when the controller calculates the control command, it first calculates the state error according to the guidance command and the flight state, then obtains the boundary function according to the designed control constraint index and the boundary adjustment law of the adaptive control index, and uses the obstacle transformation to perform an unconstrained conversion on the constrained state error variable to obtain an unconstrained auxiliary variable; further, an adaptive variable gain preset time controller and a variable gain adjustment law that can control parameters according to flight performance and flight state information are used to obtain the control command, and output it to the aircraft model, thereby completing the system closed-loop control.

Claims

1. A preset index control method based on intelligent adaptive boundary adjustment, characterized in that The method includes the following steps: Step 1: Design the basic form of the preset performance boundary function according to the task scenario and the corresponding performance requirements, and obtain the control equation under the constrained state according to the dynamic equation; Step 2: Select the state variables that can affect the control performance among the real-time flight state, deformation state and multi-task requirements, and combine the full-feedback adaptive recurrent neural network model to assign weights to the parameter values in the preset performance boundary function and the subsequent control gains, so as to achieve the purpose of adaptively adjusting the control performance; Step 3: Perform an equivalent unconstrained transformation on the constrained control model based on the obstacle function to obtain an unconstrained equivalent control model based on the auxiliary variable, so as to facilitate the design of the controller; Step 4: Preset the time stability theory and the backstepping method to design the controller, so as to complete the design of the preset index controller that can adjust the control response speed with only one parameter and can constrain the control performance, and the control gain can be set in advance according to the flight mission and its corresponding performance requirements.

2. The control method based on intelligent adaptive boundary adjustment preset index according to claim 1, characterized in that The specific steps of Step 1 are as follows: Step 1.1: Analyze the control accuracy ρ, overshoot σ, and response time T required by the control system according to the task scenario, and design the basic form of the preset performance boundary function ρ: ∞ , overshoot σ a , and response time T a , where ρ0 is the initial value of the performance function, ρ ∞ is the final value of the performance function, t is time, and τ is a constant; Step 1.2: According to the aircraft dynamic equation (2), integrate the state variables and give the aircraft state space equation (3) for the design of the surface control system: where, Ω represents the Euler angle vector; ω represents the attitude angular velocity vector; ρ1 represents the constraint on the system state quantity x1; ρ2 represents the constraint on the system state quantity x2; J represents the inertia matrix; R represents the attitude transformation matrix; d is an unknown disturbance; M is the torque vector; R, J, ω × The expressions of are as follows: where \(x_1 = \Omega\), \(x_2=\omega\); \(M\) δc represents the control command vector, \(M\) Δδ represents the moment caused by the aerodynamic rudder deviation, \(M_0\) represents the moment generated by the projectile body; \(f=-J\) -1 \(\omega\) × \(J\omega+J\) -1 \(M_0\), \(D = J\) -1 \(d+J\) -1 \(M\) Λδ represents the total disturbance; \(\sigma\), \(\beta\), \(\alpha\) are the bank angle, sideslip angle and angle of attack respectively; \(\omega\) x , \(\omega\) y , \(\omega\) z are the roll angular velocity, yaw angular velocity and pitch angular velocity respectively; Thus, the state error equation is obtained as: where e1 = x1 - x d , e2 = x2 - x 2d , x d represents the desired command for x1, x 2d represents the desired command for x2.

3. The control method based on intelligent adaptive boundary adjustment preset index according to claim 2, characterized in that The specific steps of Step 2 are as follows: Step 2.1: Select the real-time flight state Ω = [σ, β, α] T , the sweep angle Λ, the span L, the response time requirement, and the control accuracy requirement variables after the deformation state as the state variables for neural network training; Step 2.2: Construct a neural network using the radial basis function and adjust the weight vector to obtain the value of the adaptive boundary state that is affected; Among them, is the weight coefficient matrix, φ(x) represents the activation function, and ρ a (x) is the adaptive adjustment boundary term, so as to obtain the adaptive boundary adjustment function:

4. The method for controlling preset indicators based on intelligent adaptive boundary adjustment according to claim 3, wherein The specific steps of Step 3 are as follows: Select the form of the obstacle function as: where, ξ is the auxiliary error variable, and combined with the state error equation, the state error equation under unconstrained conditions is:

5. The control method based on intelligent adaptive boundary adjustment preset index according to claim 4, characterized in that The specific steps of Step 4 are as follows: Step 4.1: Design the virtual control quantity in the following form according to the first formula of Equation (11) and the preset time stability theory: where 0 < κ < 1, T(x) is the convergence time adjustment term, and x 2d0 is the equivalent virtual control term; Step 4.2: Design the control quantity in the following form according to the second formula of Equation (11): Among them, M δ0 is the equivalent control item, and thus the controller design is completed.

Images