Aircraft platform control method based on p-i inverse model and active disturbance rejection sliding mode

Abstract

The application belongs to the field of aerial remote sensing stable platform control method, and relates to an aerial platform control method based on P-I inverse model and self-disturbance sliding mode, which comprises the following steps: a P-I hysteresis nonlinear model of a piezoelectric actuator is established, and an inverse model is constructed as a feedforward controller, which is used for compensating the known hysteresis nonlinearity; a feedback controller containing an extended state observer (ESO) and a super-twisting sliding mode (STSM) is constructed, the ESO is used for estimating the lumped unknown disturbance in real time, and the feedback control quantity is calculated based on the disturbance estimation value through the STSM; finally, the feedforward and feedback control quantities are superposed to drive the stable platform. This control mode can simultaneously solve the hysteresis nonlinearity of a piezoelectric ceramic (PZT) actuator, the chattering problem of a traditional sliding mode and the suppression problem of external strong disturbance, and high-precision tracking and strong robustness of the aerial remote sensing stable platform are realized.

Description

Technical Field

[0001] This invention belongs to the field of control methods for airborne remote sensing stabilization platforms, specifically relating to an airborne platform control method based on the PI inverse model and active disturbance rejection sliding mode, used for precision servo control of airborne remote sensing stabilization platforms. Background Technology

[0002] Aerial remote sensing is a comprehensive detection technology that uses manned aircraft, unmanned aerial vehicles, hot air balloons, and other platforms to acquire ground information using technologies such as optics and radar. It is mainly used in fields such as topographic mapping, disaster prediction, and military reconnaissance.

[0003] As a crucial component of aerial remote sensing systems, aerial cameras are inevitably affected by external disturbances such as vibration during operation, resulting in poor image quality that fails to meet expectations. The image quality of aerial remote sensing cameras directly impacts the development level of remote sensing technology. Therefore, improving the image quality of remote sensing systems is currently a hot research topic in the field. Aerial remote sensing stabilization platforms are core equipment for aerial photography, reconnaissance, and surveying missions. Their main function is to isolate the interference of the movement of drones, aircraft, and other carriers, as well as external environmental vibrations (such as air disturbances), on the line of sight of the remote sensing camera, ensuring that the camera can stably and accurately align with the target, thereby acquiring high-resolution, clear images.

[0004] With the development of modern remote sensing technology, increasingly stringent requirements have been placed on the control performance of stable platforms, mainly in the following two aspects: high precision and high bandwidth, the platform needs extremely high tracking accuracy and extremely fast response speed to compensate for high-frequency vibrations; strong robustness, the aviation environment is complex, the platform must be able to strongly suppress strong vibrations from the carrier engine and uncertain airflow.

[0005] To address the aforementioned challenges, existing control methods mainly focus on PID control, traditional sliding mode control, and compensation control based on hysteresis models, but all have certain limitations.

[0006] First, while traditional PID control is simple in structure and widely used, it is essentially a linear controller. Due to the severe hysteresis and creep of piezoelectric ceramic (PZT) actuators, linear PID controllers struggle to establish accurate inverse compensation mechanisms, resulting in significantly increased tracking errors during long strokes or high-frequency motions, making it difficult to meet the requirements of high-precision imaging.

[0007] Secondly, to improve the system's anti-interference capability, sliding mode variable structure control (SMC) is widely used in airborne stabilization platforms. However, traditional first-order sliding mode control, when approaching the sliding surface, generates high-frequency "chattering" due to the presence of the switching function. This phenomenon is directly transmitted to the stabilization platform, leading to image blurring and line-of-sight jitter, severely affecting the resolution of remote sensing images.

[0008] To address the hysteresis characteristics of piezoelectric ceramic (PZT) sensors, existing technologies often employ the Prandtl-Ishlinsky (PI) model to construct an inverse model for feedforward compensation. However, this method is typically used as open-loop control, relying excessively on the model's identification accuracy. In actual aerospace conditions, once subjected to unknown external disturbances such as aircraft vibration and wind drag torque, or due to PZT aging causing model parameter mismatch, single feedforward control cannot eliminate the resulting compound errors, leading to poor system robustness.

[0009] In summary, existing single control methods are insufficient to simultaneously address the hysteresis nonlinearity of PZT, the chattering problem of traditional sliding mode, and the suppression of strong external disturbances. Therefore, a composite control method that integrates precise hysteresis compensation, chatter-free robust control, and active disturbance rejection technology is urgently needed. Summary of the Invention

[0010] In view of the shortcomings and deficiencies of the existing technology, the purpose of this invention is to provide an airborne platform control method based on the PI inverse model and active disturbance rejection sliding mode. The method includes: establishing a PI hysteresis nonlinear model of the piezoelectric actuator and constructing an inverse model as a feedforward controller to compensate for the known hysteresis nonlinearity; constructing a feedback controller including an extended state observer (ESO) and a super-spiral sliding mode (STSM), using the ESO to estimate the lumped unknown disturbance in real time, and using the STSM to calculate the feedback control quantity based on the disturbance estimate and the system tracking error; finally, superimposing the feedforward and feedback control quantities to drive the stable platform, thereby simultaneously solving the hysteresis nonlinearity of the piezoelectric ceramic (PZT) actuator, the chattering problem of the traditional sliding mode, and the suppression problem of strong external disturbances, realizing high-precision tracking and strong robustness of the airborne remote sensing stable platform.

[0011] To achieve the above objectives, the present invention adopts the following technical solution:

[0012] A control method for an aerospace platform based on the PI inverse model and active disturbance rejection sliding mode includes the following steps:

[0013] Step S1. Establish a dynamic model of the airborne remote sensing stabilization platform driven by the PZT actuator, wherein the dynamic model includes the platform angle. angular velocity Composite control voltage of PZT actuator The coupling relationship between them; and the hysteresis nonlinear characteristics of the PZT actuator are identified and modeled using the PI hysteresis nonlinear model;

[0014] Step S2. Construct the PI inverse model based on the PI hysteresis nonlinear model, and use it as a feedforward controller to calculate the feedforward control quantity;

[0015] Step S3. Convert the dynamic model in step S1 into state-space form, set the tracking error of the control system, and establish the dynamic equation of the error system;

[0016] Step S4. Construct a feedback controller comprising an extended state observer and a superspiral sliding mode; wherein, the extended state observer is used to estimate the lumped unknown disturbance of the control system in real time based on the current state, and the superspiral sliding mode calculates the feedback control quantity based on the estimated value of the lumped unknown disturbance and the tracking error of the control system, the feedback control quantity consisting of an equivalent control term and a switching control term, the equivalent control term being used to cancel the first derivative of the sliding surface. The known terms and observed disturbance terms are used to switch the control terms, which are obtained using the superhelical algorithm to ensure that the control system converges to the sliding surface.

[0017] Step S5. The feedforward control quantity and the feedback control quantity are superimposed to obtain the final composite control quantity, and then applied to the PZT actuator to enable the platform to track the desired trajectory.

[0018] As a preferred embodiment of the present invention, the PI hysteresis nonlinear model describes the input voltage of the PZT actuator by weighted superposition of multiple Play operators. With output torque The relationship between them.

[0019] As a preferred embodiment of the present invention, the feedforward control quantity is calculated by the inverse PI model, and the expression is:

[0020] ;

[0021] in, This is the feedforward control quantity. These are the weighting coefficients of each Play operator in the PI inverse model. The ideal driving torque is calculated based on the desired trajectory. For having a threshold The Play operator, The number of Play operators.

[0022] As a preferred embodiment of the present invention, step S3 converts the dynamic model in step S1 into a state-space form, and the state-space equation is expressed as:

[0023] ;

[0024] ;

[0025] From the platform perspective angular velocity , and For known parameters of the control system, if there is no stiffness term, then , , ; J The moment of inertia of the control system; C This refers to the viscous damping coefficient of the control system. and They are respectively and The first derivative, For aggregated unknown disturbances;

[0026] The tracking error of the control system includes position tracking error. and speed tracking error The established dynamic equations of the error system are as follows: ;

[0027] in, for The second derivative, For the velocity tracking error of the error system, for The second derivative, The trajectory to be tracked; From the perspective of the representative platform, This represents the platform's angular velocity.

[0028] As a preferred embodiment of the present invention, the expression of the extended state observer is:

[0029] ;

[0030] ;

[0031] ;

[0032] in, To extend the state observer gain, To provide real-time estimates of the position and velocity of the control system from an extended state observer. This is an estimate of the lumped unknown disturbance. for The first derivative, for The first derivative, for The first derivative.

[0033] As a preferred embodiment of the present invention, the first derivative of the sliding surface The expression is:

[0034] ;

[0035] Equivalent control items The expression is: ;

[0036] Switch control item The expression is: ;

[0037] in, , It is the gain of the superspiral sliding mode controller. For symbolic functions, It is a sliding surface.

[0038] The present invention also provides a control system for implementing the control method described above. The control system includes a PI inverse model, an extended state observer, and a superspiral sliding mode controller. The PI inverse model serves as a feedforward controller for calculating the feedforward control quantity. The extended state observer and the superspiral sliding mode controller together form a feedback controller. The extended state observer is used to estimate the lumped unknown disturbance of the control system in real time based on the current state. The superspiral sliding mode controller calculates the feedback control quantity based on the estimated value of the lumped unknown disturbance and the tracking error of the system.

[0039] The advantages and beneficial effects of this invention are as follows:

[0040] (1) Linearization of nonlinear systems is achieved, reducing the difficulty of controller design. This invention utilizes the PI inverse model as a feedforward controller to accurately cancel the inherent hysteresis nonlinearity of piezoelectric actuators (PZT). Through this feedforward compensation, the input-output relationship of the controlled object is macroscopically integrated into a linear system, so that the design of the subsequent feedback controller is no longer constrained by the influence of complex hysteresis loops, and the dynamic tracking capability of the system to the desired trajectory is improved.

[0041] (2) Effectively suppresses the "jitter" phenomenon of traditional sliding mode, ensuring high-resolution imaging quality. Compared with traditional first-order sliding mode control, this invention adopts the super-spiral sliding mode (STSM) algorithm. This algorithm generates a continuous and smooth control signal by transferring discontinuous sign functions to higher-order integral terms. It avoids mechanical resonance caused by high-frequency control signal switching, prevents physical damage to the precision optomechanical structure, thereby eliminating line-of-sight jitter and ensuring the imaging clarity of the airborne remote sensing camera under long exposure times.

[0042] (3) It possesses a strong active resistance capability to "lumped disturbances," reducing its dependence on the system model. This invention introduces an extended state observer (ESO) to uniformly treat the compensation residuals of the PI inverse model, system parameter perturbations (such as stiffness changes), and external environmental disturbances (such as aircraft vibration and airflow torque) as "lumped unknown disturbances" for real-time estimation and compensation. This ensures that the feedback controller can achieve high-performance control even without precisely knowing all physical parameters, improving the system's adaptability in complex and ever-changing aerospace environments.

[0043] (4) Expanding the closed-loop bandwidth of the system improves its fast response performance. This invention adopts a composite control architecture of "feedforward + feedback". The feedforward path provides the main driving energy based on the PI inverse model, enabling the system to respond quickly to high-frequency reference commands; the feedback path focuses on eliminating errors. This division of labor and cooperation mechanism breaks through the bandwidth limitation of single feedback control, enabling the stable platform to effectively isolate high-frequency carrier vibration and meet the requirements of airborne remote sensing for high-frequency line-of-sight stabilization.

[0044] (5) Enhanced system robustness to device aging and parameter drift. Piezoelectric ceramic actuators exhibit aging phenomena such as polarization decay or hysteresis drift after long-term use. The ESO in this invention can detect feedforward model mismatch caused by device aging in real time and automatically correct it as a disturbance term in the feedback loop. Therefore, the method of this invention can maintain high-precision control performance throughout the entire life cycle of the equipment without frequent re-identification of model parameters, thus reducing maintenance costs. Attached Figure Description

[0045] To further illustrate the specific embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. The drawings in this specification are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This invention provides a block diagram of the control principle of an aviation platform based on the PI inverse model and active disturbance rejection sliding mode, which shows the signal flow of the closed-loop control system and the connection relationship of each control link.

[0047] Figure 2 This is an exploded view of the control system architecture provided by the present invention, showing the composite control architecture of the system consisting of a state observation module, a compensation module, and a control module;

[0048] Figure 3 This is a flowchart of the control method provided by the present invention, which shows the complete execution steps from parameter initialization, data acquisition, state reconstruction, composite control calculation to state update;

[0049] Figure 4 The figures show a simulation comparison between the control method provided by this invention and the traditional PID control method under sinusoidal signal tracking; where (a) is a comparison of tracking performance and (b) is a comparison of tracking error. Detailed Implementation

[0050] The present invention will now be described in detail with reference to the embodiments shown in the accompanying drawings. However, these embodiments do not limit the present invention, and any structural, methodological, or functional modifications made by those skilled in the art based on these embodiments are included within the scope of protection of the present invention.

[0051] like Figures 1 to 3 As shown, this embodiment provides an airborne platform control method based on the PI inverse model and active disturbance rejection sliding mode. The method includes the following steps:

[0052] Step S1. Construct a dynamic model of the airborne remote sensing stabilization platform:

[0053] This step obtains the deterministic part of the control system by constructing a dynamic model; the indeterminate parts, such as noise, are uniformly considered as the total disturbance. The controlled object is an airborne remote sensing stabilization platform driven by a piezoelectric ceramic (PZT) actuator. The dynamic characteristics of the control system are derived from the control voltage. From the platform perspective It can be modeled as a second-order system coupled with the strong hysteresis nonlinearity of PZT; its dynamic equations can be described as:

[0054]

[0055]

[0056] in, This is the actual output angle of the airborne remote sensing stabilization platform. For the angular velocity of the airborne remote sensing stabilization platform, Angular acceleration, J Let the system's moment of inertia be denoted by . C The viscous damping coefficient of the system; For unknown external disturbances; The actual driving torque (output torque) generated by the PZT actuator. The composite control voltage (input voltage) applied to the PZT; To describe voltage and torque Prandtl-Ishlinsky (PI) hysteresis nonlinear model of the relationship between them.

[0057] In this embodiment, the PI hysteresis nonlinear model Through weighted superposition NThe basic "Play" operator (or Hysteron operator) To achieve this:

[0058]

[0059] in, The number of Play operators, ranging from 10 to 20. For having a threshold The Play operator; Weight coefficients for each Play operator; and These are the PI model parameters identified through experimental data.

[0060] Step S2. Design of the feedforward-feedback composite controller:

[0061] Based on the PI hysteresis nonlinear model, an inverse PI model is constructed as a feedforward controller to calculate the feedforward control quantity to compensate for the known hysteresis nonlinearity of PZT; a feedback controller including an extended state observer (ESO) and a super-spiral sliding mode (STSM) is constructed; wherein, the ESO is used to estimate the lumped unknown disturbance of the system in real time, and the STSM is used to calculate the feedback control quantity based on the estimated value of the lumped unknown disturbance and the system tracking error;

[0062] Specifically, in this embodiment, step S2 includes the following steps:

[0063] Step S201. Construct the PI inverse model based on the PI hysteresis nonlinear model as a feedforward controller, and calculate the feedforward control quantity to compensate for the known hysteresis nonlinearity of PZT;

[0064] Based on the PI hysteresis nonlinear model in step S1, the desired "ideal" driving torque is... It should be able to Perfectly track the expected trajectory Therefore, the ideal driving torque is determined by the desired trajectory. The derivative determines:

[0065]

[0066] in, Let the first derivative of the desired trajectory be... Let be the second derivative of the desired trajectory.

[0067] In order to generate this ideal driving torque The feedforward control quantity that needs to be applied Calculated using the PI inverse model:

[0068]

[0069] PI hysteresis nonlinear model One of its key advantages is that it has an analytic inverse. The inverse model Similarly, it can be determined by a set of Play operators. It is formed by superposition, only its weighting coefficients are added together. Unlike the PI hysteresis nonlinear model ( can be (derived from analytical calculations)

[0070]

[0071] in, These are the weighting coefficients of each Play operator in the PI inverse model. This is the ideal driving torque calculated based on the desired trajectory.

[0072] Step S202. State space transition:

[0073] To facilitate subsequent feedback controller design and analysis, the dynamic model in step S1 needs to be modified. , Physical parameters (such as those) are converted into state-space form. Let the platform angle... angular velocity and introduce and The system state-space equations can be expressed as:

[0074]

[0075]

[0076] in, This is the input voltage of the PZT actuator, i.e., the total control input applied to the PZT (feedforward control quantity + feedback control quantity). and For known parameters of the control system (if there is no stiffness term) ); The lumped unknown disturbances include the compensation residue of the PI inverse model and external environmental disturbances. And all unknowns, including unmodeled dynamics.

[0077] Step S203. Construct a feedback controller that includes an extended state observer (ESO) and a super-helical sliding mode (STSM), and calculate the feedback control quantity;

[0078] Step S2031. Establish the dynamic equations of the error system;

[0079] In this embodiment, before constructing the feedback controller, the dynamic equations of the error system are established, and the tracking error of the control system is defined as follows:

[0080]

[0081] in, For the position tracking error of the error system, This is the first derivative of the position tracking error; For the velocity tracking error of the error system; The trajectory to be tracked; From the platform's perspective, i.e., the actual location. The platform's angular velocity;

[0082] For error By differentiating the equations and substituting them into the state-space equations above, we can derive the dynamic equations of the error system as follows:

[0083]

[0084] in, For the dynamic parameters of the error system, for The second derivative of .

[0085] Step S2032. Design the Extended State Observer (ESO):

[0086] Introducing state variables For the control system state respectively and aggregated unknown disturbances Make an estimate:

[0087]

[0088] in, To extend the state observer gain, To provide real-time estimates of the position and velocity of the control system from an extended state observer. That is, the lumped unknown disturbance The estimated value, State variables First derivative, State variables First derivative, State variables First derivative.

[0089] Step S2033. Design the super-spiral sliding mode controller:

[0090] Define the sliding surface for:

[0091]

[0092] in, This is the sliding surface coefficient, with a value ranging from 50 to 200;

[0093] Differentiating equation (10) yields:

[0094] In order to make To satisfy the stability condition of the superspiral algorithm, the feedback control quantity is... Designed as an equivalent control item and switching control items The sum of the two parts:

[0095]

[0096] Among them, equivalent control items Used to offset the sliding surface The known terms and observed (ESO estimated) perturbation terms in the data. ( (estimated value)

[0097]

[0098] Switch control item A superhelical algorithm is used to ensure that the system converges to the sliding surface:

[0099]

[0100] in, , It is the gain of the superspiral sliding mode (STSM) controller. It is a symbolic function.

[0101] Finally, and By merging, the final feedback control quantity is obtained. The expression is:

[0102]

[0103] Step S3. The feedforward control quantity and the feedback control quantity are superimposed to obtain the final composite control law, and then applied to the PZT actuator to achieve precise tracking of the desired trajectory by the airborne remote sensing stabilization platform.

[0104] Specifically, in this embodiment, the composite control voltage ultimately applied to the PZT actuator... The expression is:

[0105]

[0106] in, The feedforward control quantity of the PI inverse model calculated in step S201; The feedback control quantity is determined in step S203.

[0107] This embodiment also provides an aviation platform control system based on the PI inverse model and active disturbance rejection sliding mode. Figure 1 The control principle block diagram of the air platform control method based on the PI inverse model and active disturbance rejection sliding mode is shown, as follows: Figure 1 As shown, the control system mainly includes a PI inverse model, an extended state observer (ESO), and a superspiral sliding mode controller (STSMC); among which, The desired corner trajectory of the airborne remote sensing stabilized platform; The composite control voltage applied to the piezoelectric ceramic actuator is composed of the superposition of feedforward control and feedback control. The actual rotation angle output by the system; , , These are the real-time estimates of the system's position, velocity, and total disturbance (including model uncertainties and external disturbances) from the extended state observer.

[0108] To verify the effectiveness of the proposed composite control method, a system simulation model was built in the Matlab / Simulink environment, and comparative experiments were conducted. The experimental setup is as follows: The desired trajectory was set. With an amplitude of 0.001 rad and a frequency A sinusoidal signal was used to simulate the reciprocating scanning motion of an airborne remote sensing platform. To test the system's anti-interference performance, during the simulation time... When a sudden external torque disturbance is superimposed on the system... (Simulating sudden airflow changes or aircraft maneuvers). Comparison object selection: A traditional PID controller was selected as the comparison object. This traditional controller only calculates the position error. and speed error The control voltage is generated by a linear combination of [variable name], but it does not include feedforward compensation for the hysteresis of piezoelectric ceramics using an inverse PI model, nor does it possess the ability to observe and actively compensate for external disturbances. Experimental results are analyzed as follows: Figure 4 , Figure 4 The simulation results of the controller of this invention and the traditional PID controller under the same operating conditions are shown. Figure 4 (a) Tracking performance comparison chart: showing the actual output angle of the system. For the expected trajectory The real-time tracking situation is shown. It can be seen that at the moment the disturbance is applied at 0.8s, the tracking curve of the traditional controller deviates significantly and the recovery time is long; while the controller of the present invention is almost unaffected and maintains close tracking. Figure 4(b) is a comparison graph of tracking errors, showing the changes in tracking errors for both methods. The results indicate that the composite control method of this invention keeps the tracking error within an extremely small range (on the order of magnitude of...). (rad), and when disturbances occur, the error fluctuations are quickly suppressed by relying on the fast estimation of ESO and the robustness of STSM, which verifies the advantages of this invention in terms of high accuracy and strong anti-disturbance.

[0109] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other forms without departing from the spirit or essential characteristics of the present invention.

[0110] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This way of describing the specification is only for clarity. The technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A control method for an airborne platform based on the PI inverse model and active disturbance rejection sliding mode, characterized in that, The method includes the following steps: Step S1. Establish a dynamic model of the airborne remote sensing stabilization platform driven by the PZT actuator, wherein the dynamic model includes the platform angle. angular velocity Composite control voltage of PZT actuator The coupling relationship between them; and the hysteresis nonlinear characteristics of the PZT actuator are identified and modeled using the PI hysteresis nonlinear model; Step S2. Construct the PI inverse model based on the PI hysteresis nonlinear model, and use it as a feedforward controller to calculate the feedforward control quantity. ; Step S3. Convert the dynamic model in step S1 into state-space form, set the tracking error of the control system, and establish the dynamic equation of the error system; Step S4. Construct a feedback controller comprising an extended state observer and a superspiral sliding mode; wherein, the extended state observer is used to estimate the lumped unknown disturbance of the control system in real time based on the current state, and the superspiral sliding mode calculates the feedback control quantity based on the estimated value of the lumped unknown disturbance and the tracking error of the control system, the feedback control quantity consisting of an equivalent control term and a switching control term, the equivalent control term being used to cancel the first derivative of the sliding surface. The known terms and observed disturbance terms are used to switch the control terms, which are obtained using the superhelical algorithm to ensure that the control system converges to the sliding surface. Step S5. The feedforward control quantity and the feedback control quantity are superimposed to obtain the final composite control quantity, and then applied to the PZT actuator to enable the platform to track the desired trajectory. The PI hysteresis nonlinear model describes the composite control voltage of the PZT actuator by weighted superposition of multiple Play operators. With output torque The relationship between them; The feedforward control quantity is calculated using the PI inverse model, and its expression is: ； in, This is the feedforward control variable. These are the weighting coefficients of each Play operator in the PI inverse model. The ideal driving torque is calculated based on the desired trajectory. For having a threshold The Play operator, The number of Play operators; Step S3 converts the dynamic model from step S1 into state-space form, and the state-space equations are expressed as follows: ； ； From the platform perspective angular velocity , and For known parameters of the control system, if there is no stiffness term, then , , ; J For the rotational inertia of the control system; C This refers to the viscous damping coefficient of the control system. and They are respectively and The first derivative, For aggregated unknown disturbances; The tracking error of the control system includes position tracking error. and speed tracking error The established dynamic equations of the error system are as follows: ; in, for The second derivative, For the velocity tracking error of the error system, for The second derivative, The trajectory to be tracked; From the perspective of the representative platform, Represents the platform's angular velocity.

2. The airborne platform control method based on the PI inverse model and active disturbance rejection sliding mode as described in claim 1, characterized in that, The expression for the extended state observer is: ； ； ； in, To extend the state observer gain, These are the real-time estimates of the control system's position and velocity from the extended state observer, respectively. This is an estimate of the lumped unknown disturbance. for The first derivative, for The first derivative, for The first derivative.

3. The aerospace platform control method based on the PI inverse model and active disturbance rejection sliding mode according to claim 2, characterized in that, The first derivative of the sliding surface The expression is: ； Equivalent control items The expression is: ; Switch control item The expression is: ; in, , It is the gain of the superspiral sliding mode controller. For symbolic functions, For sliding surface, This is the sliding surface coefficient.

4. An aviation platform control system based on the PI inverse model and active disturbance rejection sliding mode, characterized in that, The control system is used to implement the control method according to any one of claims 1 to 3. The control system includes a PI inverse model, an extended state observer, and a superspiral sliding mode controller. The PI inverse model serves as a feedforward controller for calculating the feedforward control quantity. The extended state observer and the superspiral sliding mode controller together form a feedback controller. The extended state observer is used to estimate the lumped unknown disturbance of the control system in real time based on the current state. The superspiral sliding mode controller calculates the feedback control quantity based on the estimated value of the lumped unknown disturbance and the tracking error of the system.

Images