Recoverable rocket attitude control method

Abstract

The application discloses a recoverable rocket attitude control method and particularly relates to the technical field of spacecraft control. In the vertical returning process of a rocket substage, the actual airflow angle and the actual three-axis rotation angular velocity are obtained through an inertial measurement combination; an affine nonlinear kinematic model is established through a nonlinear dynamic inverse outer loop controller, a dynamic deviation of the airflow angle is extracted for PID control compensation, three-axis angular velocity commands are generated through decoupling mapping in combination with an inverse model; a nonlinear spatial rotation dynamics model is established through a dynamic inverse inner loop controller, a dynamic deviation of the angular velocity is extracted for PID control compensation, three-axis expected control torques are generated through mapping in combination with the rotational inertia and the inverse model; and finally, the torque commands are distributed according to the mechanism configuration in different returning stages to drive the actions of the thrust vector, the grid fin and the attitude control nozzle.

Description

Technical Field

[0001] This invention relates to the field of spacecraft control technology, specifically to a method for attitude control of a reusable rocket. Background Technology

[0002] With the development of space transportation technology, after the first and second stages of a reusable launch vehicle separate, the first stage needs to fly back along a predetermined trajectory and land vertically in a fixed recovery area. This vertical return process faces extreme flight conditions, traversing both inside and outside the atmosphere, with severe aerodynamic interference and numerous uncertainties. At the same time, various actuators, such as engine thrust vectoring, grid fins, and attitude control nozzles, work alternately or in coordination at different stages of the return, placing extremely high demands on the control precision and response speed of the flight attitude.

[0003] Currently, attitude control in rocket flight control often employs a proportional-integral-differential superposition correction network (PIDE) based on small-disturbance linearization theory. However, during the extensive maneuvering return phase of reusable rockets, flight states change drastically, and severe cross-coupling exists between channels. Linearization models based on a single equilibrium point suffer severe mismatches in actual flight. This traditional control architecture, when faced with rapidly time-varying, highly nonlinear, and strongly coupled flight characteristics involving multiple actuators, suffers from a sharp decline in robustness due to the inability to effectively decouple channels. This leads to attitude divergence, control command oscillations, and poor adaptability to external disturbances. Therefore, it remains necessary to provide a reusable rocket attitude control method to improve attitude stability and control accuracy during the first-stage return process. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a reusable rocket attitude control method to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a reusable rocket attitude control method, comprising the following steps: S1, during the vertical return of the first stage of the reusable rocket, inertial measurement units are used to collect rocket attitude measurement data and perform flight state analysis to obtain the actual airflow angle and the actual three-axis rotational angular velocity in the rocket coordinate system at the current moment. The actual airflow angle includes the angle of attack, sideslip angle, and tilt angle; S2, the actual airflow angle is processed by a nonlinear dynamic inverse outer loop controller to establish an affine nonlinear kinematic model with the actual airflow angle as the state parameter. The dynamic deviation between the desired airflow angle and the actual airflow angle is extracted and compensated by outer loop PID control. The compensation result is then combined with the results in the affine nonlinear kinematic model. The inverse model of the outer loop nonlinear control distribution function is subjected to feedback linearization decoupling mapping to generate three-axis angular velocity commands; S3, the three-axis angular velocity commands are received through the dynamic inverse inner loop controller, and a nonlinear spatial rotational dynamics model with the actual three-axis rotational angular velocities as state parameters is established. The dynamic deviation between the three-axis angular velocity commands and the actual three-axis rotational angular velocities is extracted and compensated by inner loop PID control. Nonlinear feedback mapping is performed by combining the rocket's three-axis rotational inertia parameters and the inverse model of the nonlinear spatial rotational dynamics model to generate the desired three-axis control torque in the rocket body coordinate system; S4, according to the configuration of the control actuators in different stages of rocket return, the desired three-axis control torque is assigned as commands to drive the thrust vector, grid fins and attitude control nozzles.

[0006] In a preferred embodiment, during the vertical return of the first stage of a reusable rocket, the specific process of acquiring attitude measurement data of the rocket body and performing flight state analysis through an inertial measurement unit is as follows: using the inertial measurement unit configured inside the first stage of the reusable rocket, the original angular rate parameters and linear acceleration parameters that characterize the spatial attitude changes of the rocket body are acquired in real time at high frequency; according to the preset navigation calculation frequency, the original angular rate parameters and linear acceleration parameters are subjected to strapdown inertial navigation integration calculation and attitude matrix update analysis, and the real-time attitude data matrix of the rocket body is output.

[0007] In a preferred embodiment, the specific process of obtaining the actual airflow angle and the actual three-axis rotational angular velocity in the rocket body coordinate system at the current moment, where the actual airflow angle includes the angle of attack, sideslip angle, and tilt angle, is as follows: The absolute rotational parameters of the rocket body relative to the spatial reference base are extracted from the rocket body attitude measurement data after flight state analysis and projected onto the rocket body coordinate system to generate the actual three-axis rotational angular velocity including yaw rate, pitch rate, and roll rate; simultaneously, the wind field data and velocity vector data of the current position of the first stage of the reusable rocket are combined to calculate the angular relationship between the incoming flow and the longitudinal axis and symmetry plane of the rocket body, and the angle of attack, sideslip angle, and tilt angle corresponding to the current moment are extracted to form the actual airflow angle.

[0008] In a preferred embodiment, the actual airflow angle is processed by a nonlinear dynamic inverse outer loop controller, and an affine nonlinear kinematic model with the actual airflow angle as the state parameter is established. The specific process of extracting the dynamic deviation between the desired airflow angle and the actual airflow angle for outer loop PID control compensation is as follows: Construct a set of differential equations for the rate of change of the actual airflow angle that integrates the ballistic inclination angle and the ballistic deflection angle. Based on the set of differential equations for the rate of change of the actual airflow angle, separate and extract the nonlinear dynamic function vector independent of the control input, and the nonlinear control distribution function matrix characterizing the coupling relationship of the outer loop control, thus completing the construction of the affine nonlinear kinematic model; Simultaneously acquire the angle of attack command, sideslip angle command, and tilt angle command issued by the flight mission, compare the actual airflow angle to extract the independent tracking error of each airflow angle channel, and form a multidimensional dynamic deviation; Perform proportional, integral, and differential control law operations on the dynamic deviation in the angle of attack, sideslip angle, and tilt angle channels respectively, and output the desired airflow angle rate of change vector as the compensation result.

[0009] In a preferred embodiment, the specific process of generating a three-axis angular velocity command by combining the compensation result with the inverse model of the outer loop nonlinear control distribution function in the affine nonlinear kinematic model for feedback linearization decoupling mapping is as follows: receiving the compensation result and the nonlinear dynamic function vector in the affine nonlinear kinematic model, using the inverse model of the outer loop nonlinear control distribution function to respond to the dynamic state difference between the compensation result and the nonlinear dynamic function vector; performing inverse algebraic solution mapping based on the inverse model to eliminate the cross-coupling relationship of control variables between the angle of attack, sideslip angle and tilt angle channels, and outputting a three-axis angular velocity command acting in the rocket body coordinate system.

[0010] In a preferred embodiment, an inner loop angular velocity fast loop is constructed, and a nonlinear spatial rotational dynamics model is established using the actual three-axis rotational angular velocities as state parameters. The specific process of extracting the dynamic deviation between the three-axis angular velocity command and the actual three-axis rotational angular velocities for inner loop PID control compensation is as follows: A spatial rotational differential equation integrating the rocket's axisymmetric distribution characteristics is established. The actual three-axis rotational angular velocities are substituted into the spatial rotational differential equation as state variables to construct a nonlinear spatial rotational dynamics model reflecting the rocket's transient rotational response mechanism. The received three-axis angular velocity command and the actual three-axis rotational angular velocities are compared in each control channel to extract the rotational tracking errors in the three independent directions of pitch, yaw, and roll, constituting the rotational dynamic deviation. In the pitch, yaw, and roll channels, proportional, integral, and derivative control parameters configured in the frequency domain are introduced respectively. The inner loop PID control law is used to calculate the dynamic deviation, and the desired angular velocity change rate vector after inner loop PID control compensation is output.

[0011] In a preferred embodiment, the specific process of generating the desired three-axis control torque in the rocket body coordinate system by combining the rocket's three-axis rotational inertia parameters and the inverse model of the nonlinear spatial rotational dynamics model for nonlinear feedback mapping is as follows: The three-axis rotational inertia parameters, which include the inherent rotational properties of the rocket around the three axes in space, are obtained. The three-axis rotational inertia parameters are then mapped with the actual three-axis rotational angular velocities using a screw cross product to extract the gyro-coupled compensation term generated by the rocket's spatial rotation. The results of the inner-loop PID control compensation are then inertial-scaled using the three-axis rotational inertia parameters, and a nonlinear inverse algebraic synthesis is performed in conjunction with the gyro-coupled compensation term and the inverse model of the nonlinear spatial rotational dynamics model to restore and output the desired three-axis control torque acting in the rocket body coordinate system.

[0012] In a preferred embodiment, the specific process of distributing the desired three-axis control torque according to the control actuator configuration of different return stages of the rocket, and driving the thrust vector, grid fins, and attitude control nozzle to move is as follows: Analyze the current vertical return flight segment of the reusable rocket's first stage, and call the available state matrix and execution efficiency parameters of the thrust vector, grid fins, and attitude control nozzle within the corresponding flight segment; based on the available state matrix and execution efficiency parameters, execute the control torque distribution strategy, decomposing and mapping the desired three-axis control torque into independent underlying drive commands corresponding to each control actuator; distribute the underlying drive commands to the thrust vector deflection mechanism, grid fin servo mechanism, and attitude control nozzle valves, driving the thrust vector, grid fins, and attitude control nozzle to produce physical response actions.

[0013] In a preferred embodiment, the specific process of introducing proportional, integral, and derivative control parameters configured in the frequency domain in the pitch, yaw, and roll channels is as follows: the spatial variation characteristics of the actual airflow angle are defined as a time-cumulative slow-change state, and the spatial variation characteristics of the actual three-axis rotational angular velocity are defined as an instantaneous fast-change state; fast and slow constraint association conditions that satisfy the time-scale separation response premise are constructed; the target response frequency range of the inner loop angular velocity fast loop is set to five to ten times the target response frequency range of the outer loop airflow angle slow loop, and frequency isolation boundaries are extracted; based on the frequency isolation boundaries, the proportional, integral, and derivative control parameters in the pitch, yaw, and roll channels are independently tuned in the frequency domain, and a parameter nesting control mechanism is formed in conjunction with the outer loop airflow angle slow loop.

[0014] In a preferred embodiment, the specific process of implementing the control torque allocation strategy, which decomposes and maps the three-axis desired control torque into independent underlying drive commands corresponding to each control actuator, is as follows: When the vertical return flight segment is determined to be an extra-atmospheric region, the three-axis desired control torque is fully mapped to the attitude control nozzle valve to generate hot and cold gas injection drive commands; when the vertical return flight segment is determined to be a high-altitude three-aircraft operating range, the three-axis desired control torque is fully mapped to the thrust vector deflection mechanism to generate main engine yaw angle drive commands; when the vertical return flight segment is determined to be an aerodynamic deceleration range, the three-axis desired control torque is fully mapped to the grid rudder servo mechanism to generate control surface deflection drive commands; when the vertical return flight segment is determined to be a landing braking deceleration range, a chain-like incremental allocation mechanism is triggered, and the three-axis desired control torque is collaboratively mapped to the thrust vector deflection mechanism and the grid rudder servo mechanism according to the effectiveness weight to generate joint drive commands.

[0015] The technical effects and advantages of this invention are as follows:

[0016] (1) A reusable rocket attitude control method, wherein the actual airflow angle, including angle of attack, sideslip angle and tilt angle, is obtained in real time through inertial measurement unit during the vertical return of the first stage of the reusable rocket; an affine nonlinear kinematic model with the actual airflow angle as the state parameter is constructed using a nonlinear dynamic inverse outer loop controller, and the dynamic deviation between the desired airflow angle and the actual airflow angle is compensated by outer loop PID control, and feedback linearization decoupling mapping is performed in combination with the inverse model of the outer loop nonlinear control distribution function to generate three-axis angular velocity commands. This process abandons the small disturbance approximation assumption in the traditional control method, and directly performs nonlinear decoupling based on the full-state kinematic characteristics of the rocket, which solves the control mismatch problem caused by severe airflow angle channel coupling under large airspace maneuvering, and accurately transforms the slowly changing aerodynamic attitude adjustment into the angular velocity tracking target of the rapidly rotating rocket body.

[0017] (2) A reusable rocket attitude control method, which constructs a nonlinear spatial rotational dynamics model based on the actual three-axis rotational angular velocities using a dynamic inverse inner loop controller. After compensating for the dynamic deviation between the angular velocity command and the actual angular velocity using inner loop PID control, the desired three-axis control torque is generated by mapping the rocket's three-axis rotational inertia parameters and the inverse dynamics model. Subsequently, the control torque command is distributed to the thrust vector, grid fins, and attitude control nozzles according to the configuration of different return stages. This process constructs a double-loop nested decoupled structure with separate fast and slow loops, and cancels the spatial gyro torque coupling in the rotational dynamics through nonlinear feedback physical cancellation. This enables the system to maintain extremely high command response accuracy under strong aerodynamic disturbances and actuator coordinated switching conditions, effectively improving the robustness and landing point convergence stability of the reusable rocket attitude control.

[0018] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0019] Figure 1 This is a flowchart of a reusable rocket attitude control method according to the present invention;

[0020] Figure 2 This is a schematic diagram of the dynamic inverse method of the present invention;

[0021] Figure 3 Diagram of the fast and slow loop structure of this invention;

[0022] Figure 4 The structure diagram of the dynamic inverse control system of this invention. Detailed Implementation

[0023] This application provides an attitude control method for reusable rockets, which solves the problems of mismatch and insufficient robustness of traditional control models under the strong nonlinearity and strong coupling conditions of the vertical return phase of reusable rockets.

[0024] The overall approach of the scheme in this application embodiment is as follows: Real-time airflow angle and three-axis rotational angular velocity state parameters of the rocket are obtained using inertial measurement units; relying on a time-scale separation mechanism, the actual airflow angle is substituted into an affine nonlinear kinematic model in the outer loop, and angular velocity commands are generated through PID compensation and inverse model algebraic decoupling mapping; the actual rotational angular velocity is substituted into a nonlinear spatial rotational dynamics model in the inner loop, and the desired three-axis control torque is generated through PID compensation and inverse model dynamics decoupling mapping; finally, according to the configuration state of the actuators in the specific return phase, the torque is distributed and the underlying physical control mechanism is driven to complete the nonlinear closed-loop stable control of the rocket's flight attitude.

[0025] Example 1; please refer to Figure 1This invention provides a technical solution: a reusable rocket attitude control method, comprising the following steps: S1, during the vertical return of the first stage of the reusable rocket, inertial measurement units are used to collect rocket attitude measurement data and perform flight state analysis to obtain the actual airflow angle and the actual three-axis rotational angular velocity in the rocket coordinate system at the current moment. The actual airflow angle includes the angle of attack, sideslip angle, and tilt angle; S2, the actual airflow angle is processed by a nonlinear dynamic inverse outer loop controller to establish an affine nonlinear kinematic model with the actual airflow angle as the state parameter. The dynamic deviation between the desired airflow angle and the actual airflow angle is extracted and compensated by outer loop PID control. The compensation result is then combined with the outer loop of the affine nonlinear kinematic model. The inverse model of the nonlinear control distributed function is subjected to feedback linearization decoupling mapping to generate three-axis angular velocity commands; S3, the three-axis angular velocity commands are received through the dynamic inverse inner loop controller, a nonlinear spatial rotational dynamics model with the actual three-axis rotational angular velocities as state parameters is established, the dynamic deviation between the three-axis angular velocity commands and the actual three-axis rotational angular velocities is extracted and inner loop PID control compensation is performed, and nonlinear feedback mapping is performed by combining the rocket's three-axis rotational inertia parameters and the inverse model of the nonlinear spatial rotational dynamics model to generate the desired three-axis control torque in the rocket body coordinate system; S4, according to the configuration of the control actuators in different stages of rocket return, the desired three-axis control torque is assigned as commands to drive the thrust vector, grid fins and attitude control nozzles.

[0026] In this implementation scheme, step S1 of the rocket flight control mainly involves the low-level acquisition and analysis of flight data using an inertial measurement unit (IMU). The IMU is the core hardware for measuring the rocket's spatial attitude changes, capable of sensing the rocket's raw angular rate and linear acceleration in real time. The actual airflow angles output after analysis describe the spatial angular relationship between the rocket's attitude and the relative airflow, specifically decomposed into the angle of attack (representing the longitudinal windward angle), the sideslip angle (representing the lateral crosswind angle), and the roll angle (representing the rocket's roll state). Simultaneously, the acquired actual three-axis rotational angular velocities reflect the instantaneous rotational speed of the rocket around its three orthogonal axes in its own coordinate system. This step directly reflects the fundamental nonlinear motion of the rocket in its actual full-state state, providing accurate physical state information for subsequent feedback control.

[0027] After determining the flight state, step S2's outer-loop control is responsible for transforming the slowly changing aerodynamic attitude adjustment requirements into a rapidly rotating target tracking mechanism. Based on the significant difference between rapid and slow changes in the reusable rocket's aerodynamics, angle changes are considered as a slow loop of time accumulation. The affine nonlinear kinematic model constructed here is a structure designed for the rocket's full-state nonlinear motion equations, including a nonlinear dynamic function characterizing the spontaneous evolution of the state and an outer-loop nonlinear control distribution function reflecting the distribution of control inputs. The nonlinear dynamic inverse outer-loop controller calculates the deviation between the desired airflow angle and the current airflow angle, performs PID proportional-integral-differential calculations, and combines the result with the inverse model of the control distribution function for algebraic inversion. This process belongs to feedback linearization technology, utilizing nonlinear dynamic feedback to construct the system inverse model, eliminating the cross-coupling between the angle of attack, sideslip angle, and tilt angle, making the outer loop a normalized system with linear characteristics, ultimately outputting three-axis angular velocity commands.

[0028] After obtaining the angular velocity command, the inner loop control in step S3 is responsible for calculating the actual physical torque required to drive the rocket body to rotate. As a fast loop controlling the rotational angular rate, the inner loop typically needs a faster response speed than the outer loop to achieve time-scale separation. The nonlinear spatial rotational dynamics model established in this step incorporates the rocket's axisymmetric distribution characteristics, and the core parameter, the moment of inertia matrix, reflects the inherent inertial properties of the rocket resisting angular acceleration when rotating around its three axes. The dynamic inverse inner loop controller calculates the dynamic deviation by subtracting the command from the actual angular velocity and executes inner loop PID control. It then combines this with the moment of inertia to scale the physical dimensions and uses the inverse dynamics model for nonlinear inverse algebraic solution. This method, which combines nonlinear feedback to cancel the coupling of spatial gyroscopic torque in rotational dynamics, avoids the model mismatch problem of traditional small-disturbance linearization under strong maneuvers, and can accurately output the desired control torque acting on the three axes in the rocket body coordinate system.

[0029] In the final execution step S4, the calculated three-axis desired control torque needs to be converted into physical control actions. The vertical return of a reusable rocket's first stage to Earth involves extremely complex flight conditions, traversing both inside and outside the atmosphere. Therefore, various actuators, including thrust vectoring, grid fins, and attitude control nozzles, are configured at different return stages. Thrust vectoring generates torque by changing the direction of the main engine exhaust; grid fins provide steering force within the atmosphere using aerodynamic forces; and attitude control nozzles adjust attitude through independently controlled nozzles. This step, based on the current flight profile, invokes different numbers and orientations of actuators configured for specific stages to execute the underlying command allocation of control torque. By driving these physical components to act alternately or in combination, precise attitude convergence and stabilization control of the rocket, which exhibits rapid time-varying, highly nonlinear, and strongly coupled characteristics, is ultimately achieved.

[0030] The steps S1 to S4 above constitute the basic execution flow of the attitude control method of this invention. To more clearly reveal the underlying physical and mathematical mechanisms of the decoupling between the aforementioned dual-loop control architecture and nonlinear feedback, the control system is analyzed in depth below with reference to the accompanying schematic diagram. Please refer to... Figure 2 For flight control systems, the object is generally an affine nonlinear control system, which is described as follows: ;in, It is a state vector; For control vectors; This is the output vector; It is a nonlinear dynamic function; It is a nonlinear control distribution function; This is the output function; Figure 2 The symbol 's' appearing in the control principle block diagram represents the complex frequency domain Laplace transform operator, used to characterize the differential and integral transfer relationship of the control signal in the frequency domain. If we assume that for all... The value of , Both are reversible; the control law can be obtained by algebraically inverting the above formula, and by appropriately selecting the input. Thus, any desired dynamic model can be obtained: Further develop the rate of change of the desired state. ,use Replace the above formula This yields the final form of the dynamic inverse control law: At the physical level, this cancels out the inherent nonlinear dynamic functions of the system, thereby compensating the originally coupled nonlinear object into a pseudo-linear system with linear transfer relationships and decoupled channels. This facilitates the independent and integrated design of the external control law. Please refer to [link / reference]. Figure 3 Since the change in the attitude angle of a reusable rocket is the integral of angular velocity over time, the fast and slow variables differ significantly. Based on the principle of time-scale separation, this system is decoupled into two loops: a fast loop and a slow loop. The outer loop (slow loop) is the airflow angle control loop, with its state variables being the angle of attack, sideslip angle, and roll angle; the inner loop (fast loop) is the triaxial angular velocity control loop. To ensure the physical prerequisite of time-scale separation, the response bandwidth of the inner loop is typically configured to be 5 to 10 times that of the outer loop. The energy and motion transfer between the inner and outer loops is physically connected by the rocket's dynamics and kinematic equations. Please refer to [link to relevant documentation]. Figure 4In actual flight, the desired airflow angle command issued by the guidance system first enters the nonlinear dynamic inverse outer loop controller to calculate the desired three-axis angular velocity command. This command then serves as a reference input to the dynamic inverse inner loop controller, which, combined with the rocket's real-time rotational inertia parameters, calculates the desired three-axis control torque used to counteract gyrocoupling. Finally, through the actuator allocation logic, this desired torque is precisely allocated to the thrust vector, grid fins, and attitude control nozzles. Based on the above control system architecture and dynamic inverse decoupling mechanism, the underlying computational logic of each step will be explained in detail below, combining specific mathematical models and calculation processes.

[0031] Specifically, during the vertical return of the first stage of a reusable rocket, the process of acquiring attitude measurement data of the rocket body and analyzing its flight status through an inertial measurement unit is as follows: using the inertial measurement unit configured inside the first stage of the reusable rocket, the original angular rate parameters and linear acceleration parameters that characterize the spatial attitude changes of the rocket body are acquired in real time at high frequencies; based on the preset navigation calculation frequency, the original angular rate parameters and linear acceleration parameters are subjected to strapdown inertial navigation integration calculation and attitude matrix update analysis, and the real-time attitude data matrix of the rocket body is output.

[0032] In this implementation scheme, an inertial measurement unit (IMU) configured inside the first stage of the rocket serves as the core sensing reference. Under the complex conditions of vertical return, the sensing units within the IMU capture and output the high-frequency raw angular rate and linear acceleration parameters of the rocket body in real time. The core calculation in this step is performed through high-frequency integration and state updates. Based on the preset navigation calculation frequency, the system continuously integrates and accumulates the acquired raw parameters. Within the calculation cycle, the system recursively updates the rocket body attitude quaternion or direction cosine matrix by discrete integration of angular rate and acceleration. Through this high-frequency integration and matrix analysis, the system can effectively filter out external high-frequency noise and correct the attitude reference, ultimately updating and outputting an accurate real-time attitude data matrix for the rocket body, providing a stable and reliable physical space foundation for subsequent dual-loop nonlinear decoupled control.

[0033] Specifically, the process of obtaining the actual airflow angle and the actual three-axis rotational angular velocity in the rocket body coordinate system at the current moment, including the angle of attack, sideslip angle, and tilt angle, is as follows: Extract the absolute rotational parameters of the rocket body relative to the spatial reference base from the rocket body attitude measurement data after flight state analysis, project them onto the rocket body coordinate system, and generate the actual three-axis rotational angular velocity including yaw rate, pitch rate, and roll rate; Simultaneously combine the wind field data and velocity vector data of the current position of the first stage of the reusable rocket, calculate the angle relationship between the incoming flow and the longitudinal axis and symmetry plane of the rocket body, and extract the angle of attack, sideslip angle, and tilt angle corresponding to the current moment to form the actual airflow angle.

[0034] In this implementation scheme, the system needs to further extract the independent state parameters required for the internal and external control loops from the aforementioned attitude data matrix. The system first performs a spatial coordinate transformation, accurately projecting the absolute rotational parameters of the rocket body into its own coordinate system, thereby directly separating the yaw rate, pitch rate, and roll rate necessary for fast loop control. For obtaining the actual airflow angle, the system primarily relies on the principle of spatial geometric projection for analysis. The system simultaneously combines the rocket's real-time velocity vector with the wind field data of the surrounding airspace, determining the absolute direction of the relative incoming flow through three-dimensional vector synthesis; subsequently, it calculates the spatial geometric angle between this relative incoming flow and the rocket's longitudinal axis and the plane of symmetry. Based on the above spatial geometric mapping relationship, the angle of attack reflecting longitudinal aerodynamic force deviation, the sideslip angle reflecting lateral force deviation, and the roll angle characterizing the roll attitude are extracted. These three angles together constitute the actual airflow angle and are directly input into the external loop nonlinear kinematic model as basic state feedback parameters.

[0035] Specifically, the actual airflow angle is processed by a nonlinear dynamic inverse outer loop controller. An affine nonlinear kinematic model with the actual airflow angle as the state parameter is established. The specific process of extracting the dynamic deviation between the desired airflow angle and the actual airflow angle for outer loop PID control compensation is as follows: Construct a set of differential equations for the rate of change of the actual airflow angle that integrates the ballistic inclination angle and the ballistic deflection angle. Based on the set of differential equations for the rate of change of the actual airflow angle, separate and extract the nonlinear dynamic function vector independent of the control input, and the nonlinear control distribution function matrix characterizing the coupling relationship of the outer loop control, thus completing the construction of the affine nonlinear kinematic model. Simultaneously acquire the angle of attack command, sideslip angle command, and tilt angle command issued by the flight mission, compare the actual airflow angle to extract the independent tracking error of each airflow angle channel, and form a multidimensional dynamic deviation. Perform proportional, integral, and differential control law operations on the dynamic deviation in the angle of attack, sideslip angle, and tilt angle channels respectively, and output the desired airflow angle rate of change vector as the compensation result.

[0036] In this implementation scheme, the system performs kinematic modeling and error compensation for the slow-changing state of the outer loop based on the time-scale separation characteristic. First, it integrates the ballistic inclination angle and ballistic deflection angle to construct a set of differential equations for the actual airflow angle change rate. This separates the nonlinear dynamic function vector independent of the system control quantity and the nonlinear control distributed function matrix characterizing the control coupling relationship, establishing an affine nonlinear kinematic model. The slow loop is the outer loop controlling the airflow attitude angle. The design process of the nonlinear dynamic inverse outer loop controller is as follows: Angle of attack Sideslip angle and tilt angle The rate of change is expressed as follows: ;in, For ballistic inclination angle, For ballistic deflection, Angular velocity about the x-axis, Angular velocity about the y-axis, The angular velocity about the z-axis; For the angle of attack, Sideslip angle, The tilt angle, For the trajectory inclination angle, Let be the ballistic deflection angle. The above formula can be rearranged into the following form: ;in, For a nonlinear dynamic function vector, its three components They are respectively , , Three-channel nonlinear dynamic function; It is a nonlinear control distribution function. ; ;because However, the rocket does not exist under the control of the controller. Therefore, matrix G is always invertible. Subsequently, the system acquires guidance commands and calculates multidimensional dynamic deviations. To achieve tracking of the airflow angle, let: ;in Angle of attack command, This is a sideslip angle command. The tilt angle command is a vector composed of angular deviations. Its three components They are respectively , , Three-channel angular deviation. Using PID control, the following channel-specific control can be obtained: ;in, The angle-of-attack channel control quantity represents the desired rate of change of angle of attack. The sideslip angle channel control quantity represents the desired rate of change of the sideslip angle. Let the tilt angle channel control quantity represent the desired rate of change of tilt angle. These are the control parameters for PID control. The specific values ​​of these control parameters are obtained through comprehensive tuning using frequency domain analysis, based on the desired closed-loop response bandwidth and stability margin boundary of the control system.

[0037] Specifically, the process of combining the compensation results with the inverse model of the outer loop nonlinear control distribution function in the affine nonlinear kinematic model to perform feedback linearization decoupling mapping and generate three-axis angular velocity commands is as follows: receiving the compensation results and the nonlinear dynamic function vector in the affine nonlinear kinematic model, using the inverse model of the outer loop nonlinear control distribution function to respond to the dynamic state difference between the compensation results and the nonlinear dynamic function vector; performing inverse algebraic solution mapping based on the inverse model to eliminate the cross-coupling relationship of control variables between the angle of attack, sideslip angle, and tilt angle channels, and outputting three-axis angular velocity commands acting in the rocket body coordinate system.

[0038] In this implementation scheme, the system inputs the compensation results generated in the preceding steps into the feedback decoupling module, and uses a dynamic inverse algorithm to eliminate the severe physical cross-coupling between the various airflow angle channels. Since the rocket will not exhibit an extreme runaway attitude with a sideslip angle reaching 90 degrees under the control of the controller, the determinant of the nonlinear control distribution function matrix is ​​always non-zero, possessing the mathematical condition of global invertibility. The system receives the expected airflow angle change rate vector containing the compensation results of each channel, and combines it with the nonlinear dynamic function vector to perform inverse algebraic solution mapping. The core calculation formula is: assuming the input vector of the outer loop dynamic inverse is... This represents the desired rate of change of the airflow angle. Substituting this into the rocket's outer loop equation... By performing the inverse solution, we can obtain: Among them, by multiplying by the inverse matrix system, the originally nonlinear and mutually interfering aerodynamic coupling terms can be completely canceled at the control level, and the desired airflow angle change rate can be accurately and independently converted into a three-axis angular velocity command acting in the rocket body coordinate system.

[0039] Specifically, the process of constructing an inner-loop angular velocity fast loop, using the actual three-axis rotational angular velocities as state parameters to establish a nonlinear spatial rotational dynamics model, and extracting the dynamic deviation between the three-axis angular velocity command and the actual three-axis rotational angular velocities for inner-loop PID control compensation is as follows: A spatial rotational differential equation integrating the rocket's axisymmetric distribution characteristics is established. The actual three-axis rotational angular velocities are substituted into the spatial rotational differential equation as state variables to construct a nonlinear spatial rotational dynamics model reflecting the rocket's transient rotational response mechanism. The received three-axis angular velocity command and the actual three-axis rotational angular velocities are compared across each control channel to extract the rotational tracking errors in the three independent directions of pitch, yaw, and roll, constituting the rotational dynamic deviation. In the pitch, yaw, and roll channels, proportional, integral, and derivative control parameters configured in the frequency domain are introduced respectively. The inner-loop PID control law is then applied to the dynamic deviation, outputting the desired angular velocity rate of change vector after inner-loop PID control compensation.

[0040] In this implementation scheme, the system uses a dynamic inverse method to convert the desired rate of change of angular velocity into the actual physical driving torque. Since the rocket exhibits axisymmetric distribution characteristics, the system first obtains a simplified diagonal matrix form of the moment of inertia. To achieve tracking of the rocket's angular velocity, let: ;in A vector composed of three-axis angular velocity commands , A vector composed of the angular velocity deviations of the three axes Using PID control, the following channel-specific control can be obtained: ;in, For PID control parameters, , , The inner loop three-channel control quantity represents the desired rate of change of angular velocity; These are the desired angular velocity command components for the roll, yaw, and pitch channels, respectively. These are the actual rotation tracking errors extracted from the roll, yaw, and pitch channels, respectively.

[0041] Specifically, the process of generating the desired three-axis control torque in the rocket body coordinate system by combining the rocket's three-axis rotational inertia parameters and the inverse model of the nonlinear spatial rotational dynamics model through nonlinear feedback mapping is as follows: The three-axis rotational inertia parameters, which include the rocket's inherent rotational properties around the three axes in space, are obtained. The three-axis rotational inertia parameters are then mapped with the actual three-axis rotational angular velocities using a screw cross product to extract the gyro-coupled compensation term associated with the rocket's spatial rotation. The results of the inner-loop PID control compensation are then inertial-scaled using the three-axis rotational inertia parameters. Finally, a nonlinear inverse algebraic synthesis is performed, combining the gyro-coupled compensation term and the inverse model of the nonlinear spatial rotational dynamics model, to restore and output the desired three-axis control torque acting in the rocket body coordinate system.

[0042] In this implementation scheme, the system utilizes a dynamic inverse method to convert the desired rate of change of angular velocity into the actual physical driving torque. Due to the axisymmetric distribution characteristics of the rocket, the system first obtains the simplified diagonal matrix form of the moment of inertia. The differential equation of the angular velocity of a rotating object in space can be expressed as: ;in Here is the rotational inertia matrix. A vector composed of the angular velocities of the three axes , These represent the three-axis angular velocities in the rocket's coordinate system. Let be the torque vector in the arrow body coordinate system, where This represents the three-axis torques in the rocket's coordinate system. The equations of motion for the rocket's rotation are as follows: Rockets are axially symmetric, therefore Here is the inertia matrix; where These are the moments of inertia about the x, y, and z axes, respectively. The equations of motion for rocket rotation can be written separately for each channel: Assume the input of the inner loop dynamic inverse is It represents the rate of change of the desired angular velocity, replacing... By performing the inverse calculation, the desired torque of the rocket can be obtained: Specifically, at the physical level, the desired three-axis control torque is directly reproduced, which is completely decoupled and available for the actuator to respond to.

[0043] Specifically, based on the configuration of the control actuators at different stages of rocket reentry, the process of distributing the desired three-axis control torque to drive the thrust vector, grid fins, and attitude control nozzles is as follows: The current vertical reentry flight segment of the reusable rocket's first stage is analyzed, and the available state matrices and execution efficiency parameters of the thrust vector, grid fins, and attitude control nozzles within the corresponding flight segment are retrieved. Based on the available state matrices and execution efficiency parameters, a control torque distribution strategy is executed, decomposing and mapping the desired three-axis control torque into independent underlying drive commands corresponding to each control actuator. These underlying drive commands are distributed to the thrust vector deflection mechanism, grid fin servo mechanism, and attitude control nozzle valves, driving the thrust vector, grid fins, and attitude control nozzles to generate physical response actions. Ultimately, the dynamic correction and stabilization control of the rocket's first stage attitude is achieved through the physical response actions of these three types of hardware.

[0044] The control torque distribution strategy is as follows: Due to the drastic changes in atmospheric dynamic pressure environment at different altitudes during vertical reentry, the system first analyzes the current vertical reentry flight segment in real time. Outside the atmosphere, all control torques are evenly distributed to the attitude control nozzle (RCS). During the high-altitude three-engine operation phase of the reentry phase, all control torques are evenly distributed to thrust vector control. During the aerodynamic deceleration phase, all control torques are evenly distributed to the grid fins for aerodynamic control. During the braking and deceleration phase before landing, a chain-incremental strategy is used to evenly distribute control torques to thrust vector and grid fin control.

[0045] In a preferred embodiment, the specific process of introducing proportional, integral, and derivative control parameters configured in the frequency domain in the pitch, yaw, and roll channels is as follows: the spatial variation characteristics of the actual airflow angle are defined as a time-cumulative slow-change state, and the spatial variation characteristics of the actual three-axis rotational angular velocity are defined as an instantaneous fast-change state; fast and slow constraint association conditions that satisfy the time-scale separation response premise are constructed; the target response frequency range of the inner loop angular velocity fast loop is set to five to ten times the target response frequency range of the outer loop airflow angle slow loop, and frequency isolation boundaries are extracted; based on the frequency isolation boundaries, the proportional, integral, and derivative control parameters in the pitch, yaw, and roll channels are independently tuned in the frequency domain, and a parameter nesting control mechanism is formed in conjunction with the outer loop airflow angle slow loop.

[0046] In this implementation scheme, the system performs frequency domain decoupling tuning of the control parameters for the inner and outer loops based on a time-scale separation physical mechanism. Since changes in rocket angle and attitude depend on the integral accumulation of angular velocity over time, spatial variations in the airflow angle are defined as a slow-changing state, while rotational angular velocity is defined as a fast-changing state. To ensure that outer loop commands can be tracked by the inner loop without hysteresis, the system forcibly defines frequency isolation boundaries during the frequency domain design phase, setting the target crossing frequency and closed-loop bandwidth of the inner loop's fast angular velocity loop to five to ten times that of the outer loop's slow airflow angle loop. Under this frequency constraint, the proportional gain, integral time constant, and derivative time constant of the pitch, yaw, and roll channels are independently calculated and tuned. This nested parameter configuration mechanism cuts off dynamic interference between the inner and outer loops in the control frequency band, preventing low-frequency oscillations during attitude tracking.

[0047] Specifically, the process of implementing the control torque allocation strategy, which decomposes and maps the three-axis desired control torque into independent underlying drive commands corresponding to each control actuator, is as follows: When the vertical return flight segment is determined to be an extra-atmospheric region, the three-axis desired control torque is fully mapped to the attitude control nozzle valve to generate hot and cold gas injection drive commands; when the vertical return flight segment is determined to be a high-altitude three-aircraft operating range, the three-axis desired control torque is fully mapped to the thrust vector deflection mechanism to generate main engine yaw angle drive commands; when the vertical return flight segment is determined to be an aerodynamic deceleration range, the three-axis desired control torque is fully mapped to the grid rudder servo mechanism to generate control surface deflection drive commands; when the vertical return flight segment is determined to be a landing braking deceleration range, a chain-like incremental allocation mechanism is triggered, and the three-axis desired control torque is collaboratively mapped to the thrust vector deflection mechanism and the grid rudder servo mechanism according to the effectiveness weight to generate joint drive commands.

[0048] In this implementation scheme, to address the physical constraint of nonlinear and dramatic changes in dynamic pressure during the rocket's vertical return profile at tens of kilometers, the system implements a segmented, customized, and directional torque distribution mechanism based on flight altitude and Mach number. Specifically, when the rocket is outside the atmosphere or in an extremely thin atmosphere, the aerodynamic control surfaces are completely ineffective, and the system directly resolves the desired three-axis control torque into pulse width modulation (PWM) opening commands for the attitude control nozzle valves, implementing pure jet reverse thrust attitude adjustment. When entering the high-altitude three-engine operating range, the main engine thrust dominates, and the torque is fully mapped into displacement commands for the electro-hydraulic servo actuators of the thrust vector deflection mechanism. When entering the aerodynamic deceleration range of the dense atmosphere, to save propellant, the physical condition of a sharp increase in dynamic pressure is utilized, and the torque is fully mapped into control surface deflection commands for the grid fin servo mechanism. When reaching the landing braking deceleration range, the sudden drop in flight speed leads to a decrease in the aerodynamic effectiveness of the grid fins, triggering a chain-like incremental distribution strategy. Based on thrust vector control, and supplemented by grid fins as high-frequency fine-tuning cooperative mapping, a joint drive command is generated to ensure attitude convergence at the final landing point.

[0049] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0050] This invention is described with reference to flowchart illustrations and / or block diagrams of systems, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0051] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0052] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0053] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0054] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for attitude control of a reusable rocket, characterized in that, Includes the following steps: S1. During the vertical return of the first stage of the reusable rocket, the attitude measurement data of the rocket body is collected by the inertial measurement unit and the flight state is analyzed to obtain the actual airflow angle of the rocket at the current moment and the actual three-axis rotational angular velocity in the rocket body coordinate system. The actual airflow angle includes the angle of attack, sideslip angle and tilt angle. S2. The actual airflow angle is processed by a nonlinear dynamic inverse outer loop controller. An affine nonlinear kinematic model with the actual airflow angle as the state parameter is established. The dynamic deviation between the desired airflow angle and the actual airflow angle is extracted and compensated by outer loop PID control. The compensation result is combined with the inverse model of the outer loop nonlinear control distribution function in the affine nonlinear kinematic model to perform feedback linearization decoupling mapping and generate a three-axis angular velocity command. S3. Receive the three-axis angular velocity command through the dynamic inverse inner loop controller, establish a nonlinear spatial rotation dynamics model with the actual three-axis rotation angular velocity as the state parameter, extract the dynamic deviation between the three-axis angular velocity command and the actual three-axis rotation angular velocity for inner loop PID control compensation, and perform nonlinear feedback mapping by combining the rocket's three-axis rotation inertia parameters and the inverse model of the nonlinear spatial rotation dynamics model to generate the desired three-axis control torque in the rocket body coordinate system. S4. Based on the configuration of the control actuators for different stages of rocket return, the desired control torque of the three axes is assigned to drive the thrust vector, grid fins and attitude control nozzles.

2. The reusable rocket attitude control method according to claim 1, characterized in that: During the vertical return of the first stage of a reusable rocket, the specific process of acquiring attitude measurement data of the rocket body and performing flight state analysis through inertial measurement unit is as follows: By utilizing the inertial measurement unit configured inside the first stage of a reusable rocket, the original angular rate parameters and linear acceleration parameters that characterize the spatial attitude changes of the rocket body can be acquired in real time at high frequency. Based on the preset navigation calculation frequency, the original angular rate parameter and linear acceleration parameter are subjected to strapdown inertial navigation integral calculation and attitude matrix update analysis, and the real-time attitude data matrix of the rocket body is output.

3. The reusable rocket attitude control method according to claim 1, characterized in that: The specific process for obtaining the rocket's actual airflow angle and actual three-axis rotational angular velocity in the rocket body coordinate system at the current moment, including the actual airflow angle, angle of attack, sideslip angle, and roll angle, is as follows: The absolute rotation parameters of the rocket body relative to the spatial reference are extracted from the rocket body attitude measurement data after flight status analysis, and projected onto the rocket body coordinate system to generate the actual three-axis rotation angular velocities including yaw rate, pitch rate and roll rate. By simultaneously combining the wind field data and velocity vector data of the current position of the first stage of the reusable rocket, the angle relationship between the incoming flow and the longitudinal axis and symmetry plane of the rocket body is calculated, and the angle of attack, sideslip angle and tilt angle corresponding to the current moment are extracted to form the actual airflow angle.

4. The attitude control method for a reusable rocket according to claim 1, characterized in that: The actual airflow angle is processed by a nonlinear dynamic inverse outer loop controller. An affine nonlinear kinematic model with the actual airflow angle as the state parameter is established. The specific process of extracting the dynamic deviation between the desired airflow angle and the actual airflow angle for outer loop PID control compensation is as follows: A set of differential equations for the actual airflow angle change rate that integrates the ballistic inclination angle and the ballistic deflection angle is constructed. Based on the set of differential equations for the actual airflow angle change rate, nonlinear dynamic function vectors independent of the control input are extracted, as well as a nonlinear control distribution function matrix that characterizes the coupling relationship of the external loop control, and the affine nonlinear kinematic model is constructed. Simultaneously acquire angle of attack, sideslip, and roll commands issued by the flight mission, compare them with the actual airflow angle to extract the independent tracking error of each airflow angle channel, and form a multi-dimensional dynamic deviation; The proportional, integral, and derivative control laws are calculated for the dynamic deviation in the angle of attack, sideslip, and tilt angle channels, respectively, and the output is the vector of the expected airflow angle change rate as the compensation result.

5. The attitude control method for a reusable rocket according to claim 1, characterized in that: The specific process of generating three-axis angular velocity commands by combining the compensation results with the inverse model of the outer-loop nonlinear control distribution function in the affine nonlinear kinematic model for feedback linearization decoupling mapping is as follows: The compensation result and the nonlinear dynamic function vector in the affine nonlinear kinematic model are received, and the dynamic state difference between the compensation result and the nonlinear dynamic function vector is utilized by the inverse model response of the outer loop nonlinear control distribution function. The inverse model is used to perform inverse algebraic solution mapping to eliminate the cross-coupling relationship of control variables between the angle of attack, sideslip angle and tilt angle channels, and output the three-axis angular velocity command acting in the rocket body coordinate system.

6. The attitude control method for a reusable rocket according to claim 1, characterized in that: The specific process of constructing an inner-loop angular velocity fast loop, establishing a nonlinear spatial rotational dynamics model using the actual three-axis rotational angular velocities as state parameters, and extracting the dynamic deviation between the three-axis angular velocity commands and the actual three-axis rotational angular velocities for inner-loop PID control compensation is as follows: A spatial rotation differential equation integrating the axisymmetric distribution characteristics of the rocket is established. The actual three-axis rotational angular velocity is substituted into the spatial rotation differential equation as a state variable to construct a nonlinear spatial rotation dynamic model that reflects the transient rotational response mechanism of the rocket body. The received three-axis angular velocity commands are compared with the actual three-axis rotational angular velocities in each control channel, and the rotation tracking errors in the three independent directions of pitch, yaw and roll are extracted to form the rotational dynamic deviation. In the pitch, yaw, and roll channels, proportional, integral, and derivative control parameters configured in the frequency domain are introduced respectively. The inner-loop PID control law is used to calculate the dynamic deviation and output the desired angular velocity rate of change vector after inner-loop PID control compensation.

7. The attitude control method for a reusable rocket according to claim 1, characterized in that: The specific process of generating the desired three-axis control torque in the rocket body coordinate system by combining the rocket's three-axis rotational inertia parameters and the inverse model of the nonlinear spatial rotational dynamics model through nonlinear feedback mapping is as follows: Obtain the three-axis rotational inertia parameters that include the rocket's inherent three-axis rotational properties around space. Perform a screw cross product mapping operation on the three-axis rotational inertia parameters and the actual three-axis rotational angular velocities to extract the gyro-coupled compensation term generated by the rocket's spatial rotation. The results of the inner-loop PID control compensation are inertial scaling and mapping are performed using the three-axis rotational inertia parameters. Then, nonlinear inverse algebraic synthesis is performed by combining the gyro coupling compensation term and the inverse model of the nonlinear spatial rotational dynamics model to restore the desired three-axis control torque acting on the rocket body coordinate system.

8. The attitude control method for a reusable rocket according to claim 1, characterized in that: Based on the configuration of the control actuators for different stages of rocket reentry, the three-axis desired control torque is assigned as commands, and the specific process of driving the thrust vector, grid fins, and attitude control nozzles is as follows: Analyze the current vertical return flight segment of the first stage of the reusable rocket, and call up the available state matrix and execution performance parameters of the thrust vector, grid fins and attitude control nozzles in the corresponding flight segment; Based on the available state matrix and execution performance parameters, a control torque allocation strategy is implemented, decomposing and mapping the desired control torque of the three axes into independent underlying drive commands corresponding to each control actuator; The underlying drive commands are distributed to the thrust vector deflection mechanism, grid rudder servo mechanism, and attitude control nozzle valve, driving the thrust vector, grid rudder, and attitude control nozzle to produce physical response actions.

9. The attitude control method for a reusable rocket according to claim 6, characterized in that the specific process of introducing proportional, integral, and derivative control parameters configured in the frequency domain in the pitch, yaw, and roll channels is as follows: The spatial variation characteristics of the actual airflow angle are defined as a slow-change state with time accumulation, and the spatial variation characteristics of the actual three-axis rotational angular velocity are defined as a fast-change state with instantaneous movement. Fast and slow constraint correlation conditions that satisfy the premise of time-scaled separation response are constructed. The target response frequency range of the inner ring fast angular velocity loop is set to be five to ten times that of the target response frequency range of the outer ring slow airflow angular velocity loop, and the frequency isolation boundary is extracted. Based on the frequency isolation boundary, the proportional, integral, and derivative control parameters in the pitch, yaw, and roll channels are independently tuned in the frequency domain, forming a parameter nesting control mechanism in coordination with the outer loop airflow angle slow loop.

10. The reusable rocket attitude control method according to claim 8, characterized in that the specific process of executing the control torque distribution strategy and decomposing and mapping the three-axis desired control torque into independent underlying drive commands corresponding to each control actuator is as follows: When the vertical return flight segment is determined to be an extra-atmospheric region, the three-axis desired control torque is fully mapped to the attitude control nozzle valve to generate cold and hot gas injection drive commands; When the vertical return flight segment is determined to be the high-altitude three-aircraft working zone, the full amount of the desired control torque of the three axes is mapped to the thrust vector deflection mechanism to generate the main engine yaw angle drive command; When the vertical return flight segment is determined to be an aerodynamic deceleration zone, the full amount of the three-axis desired control torque is mapped to the grid servo mechanism to generate the control surface deflection drive command. When the vertical return flight segment is determined to be the landing braking deceleration range, a chain-incremental allocation mechanism is triggered, which coordinates the desired control torque of the three axes to the thrust vector deflection mechanism and the grid servo mechanism according to the effectiveness weight to generate joint drive commands.

Images