A fixed-time robust direct thrust control method and system for a linear model of an aero-engine
By combining a fixed-time thrust observer with a direct thrust controller, the real-time and accuracy issues in aero-engine thrust estimation are solved, fixed-time convergence and robustness of thrust tracking are achieved, computational complexity is reduced, and the limitations of traditional control methods are overcome.
Patent Information
- Application Number
- CN202511638374.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-10
- Publication Date
- 2026-11-17
AI Technical Summary
Existing direct thrust control methods for aero-engines struggle to balance real-time performance, interpretability, and engineering robustness, particularly in thrust estimation where computational complexity and accuracy are both high.
A fixed-time thrust observer module is used to construct a linear state-space model of the aero-engine through the synergistic effect of linear and nonlinear gains, so as to achieve convergence of thrust estimation error within a fixed time. Combined with the direct thrust controller, a fuel flow regulation signal is generated to ensure thrust tracking accuracy.
It achieves fixed-time convergence of thrust tracking while ensuring L2 gain performance, reduces computational resource requirements, overcomes the thrust-speed mismatch problem in traditional indirect thrust control, and improves the accuracy and robustness of thrust tracking.
Smart Images

Figure CN123113306A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of direct thrust control technology for aero-engines, and more particularly to a fixed-time robust direct thrust control method and system for a linear model of an aero-engine. Background Technology
[0002] As the core power plant of aircraft, the thrust generation and regulation of aero engines directly impact aircraft performance, fuel economy, and flight safety. In traditional control architectures, engine speed is typically the primary controlled variable, and thrust management is indirectly achieved by adjusting fuel flow to maintain stable speed. Although this architecture has been widely adopted over decades, its inherent limitations are becoming increasingly apparent in high-performance aircraft propulsion systems. Thrust and engine speed have a nonlinear coupling relationship and are influenced by various factors such as flight altitude, Mach number, and intake conditions. To ensure safe operation, traditional controllers need to introduce anti-surge margin buffers, leading to frequent suboptimal engine operation. Furthermore, as performance deteriorates (e.g., compressor blade erosion) causing changes in aerodynamic characteristics, fixed-model-based speed control struggles to achieve online optimization.
[0003] To overcome the aforementioned limitations, direct thrust control abandons engine speed as an intermediary and instead establishes a direct mapping from throttle position to thrust, thereby achieving more precise thrust adjustment. In recent years, research on direct thrust control has mainly focused on two technical approaches: model-driven and data-driven.
[0004] In model-driven approaches, existing solutions typically construct state-space models based on the engine's thermodynamic and aerodynamic characteristics and design control laws accordingly. A typical approach combines thrust observer control with a method designed to suppress parameter uncertainties and external disturbances, thereby ensuring thrust tracking accuracy. Building upon this, further developed multisimplex extension techniques, by explicitly defining key parameter boundaries and embedding an automatic scheduling mechanism, can effectively extend the controller's adaptability within a larger flight envelope while maintaining closed-loop stability and real-time performance. Other research employs explicit model predictive control, achieving high-precision thrust regulation under multivariable constraints through online optimization of time-varying models with reduced-order linear parameters; while the μ-synthetic control scheme, with its strong robustness, significantly reduces the impact of external disturbances on thrust tracking. Furthermore, gain-scheduled PID controllers can adjust parameters according to real-time flight conditions, ensuring smooth transition and stable operation of thrust across the entire flight envelope.
[0005] In terms of data-driven approaches, existing solutions largely rely on machine learning and deep learning technologies. For example, multi-layer long short-term memory networks are used to capture the nonlinear dynamics of the system based on sensor signals such as fuel flow, rotor speed, and nozzle opening, achieving high-precision online thrust estimation. Hybrid architectures also combine neural networks with linear model predictive control, using neural networks to compensate for model errors and model predictive control to handle constrained optimization problems, thus balancing thrust tracking performance and computational efficiency. Another approach borrows from the minimum model adaptive control concept, introducing an online neural network update mechanism to improve adaptability to engine performance degradation and sudden disturbances.
[0006] Overall, in safety-critical scenarios, physical model-based control methods still hold an important position due to their rigorous stability proofs and built-in safety margins; while data-driven methods show significant potential in handling high-order nonlinearities and unmodeled dynamics, and are particularly suitable for tasks requiring real-time adaptation and state estimation.
[0007] It should be noted that since thrust cannot be directly measured, accurate thrust estimation is always a prerequisite for any effective direct thrust control scheme. Therefore, both model-driven and data-driven methods must integrate or rely on specialized thrust estimation techniques. In the field of aero-engine thrust estimation, two main types of methods have emerged: model-based Kalman filter estimation and data-based "black box" models. Model-driven Kalman filter methods originated from early work on linear Kalman filters that required offline identification. Subsequent hybrid structures introduced component-level physical models at their core and coupled incremental linear Kalman filters with real-time linearization adaptive modules, thereby effectively mitigating errors caused by model mismatch. The emergence of extended Kalman filters reduced the reliance on offline modeling, but the computation of high-dimensional Jacobian matrices often leads to divergence problems. In recent years, unscented Kalman filters have been widely adopted due to their elimination of Jacobian computation and higher estimation accuracy, although estimation jumps may still occur during rapid engine transients. On the other hand, learning-based thrust observers mainly rely on neural networks and often combine them with metaheuristic optimization. Typical examples include hybrid observers employing particle swarm optimization and radial basis function networks, highly robust methods considering multi-source disturbances, and performance parameter learners implemented within deep neural network frameworks. These methods excel at capturing higher-order nonlinearities, but typically rely on large amounts of flight data and struggle to adapt to performance degradation throughout the engine's lifespan.
[0008] In summary, existing methods struggle to balance real-time performance, interpretability, and engineering robustness in the face of the computational burden of offline modeling and high-dimensional filtering, or the demands of big data and complex network structures. Therefore, a control strategy that guarantees fast convergence while possessing low computational complexity, high interpretability, and strong robustness is urgently needed. Since the introduction of fixed-time stability theory, its significant characteristic has been the ability to provide an adjustable upper bound on the convergence time, independent of initial conditions. This effectively solves the convergence delay problem of traditional finite-time control in engineering applications. As an important means of putting this theory into practice, fixed-time control has sparked continuous research interest in electromechanical systems, industrial systems, and high-dynamic aircraft control. In the field of propulsion, fixed-time control technology has begun to demonstrate clear benefits: for example, fixed-time control strategies used to suppress aero-engine surge significantly reduce aerodynamic instability and extend service life; in spacecraft control, existing adaptive fixed-time control algorithms achieve rapid obstacle avoidance and accurate trajectory tracking; to address spacecraft parameter uncertainties and external disturbances, existing technologies have designed fixed-time observers to ensure that tracking errors converge within a predetermined time and meet performance specifications; in the field of hypersonic vehicles, fixed-time fuzzy disturbance observers enable rapid estimation and compensation of aerodynamic disturbances within a specified convergence time. However, to the inventors' knowledge, fixed-time observers have not yet been systematically applied to aero-engine thrust estimation, a critical problem with stringent requirements for both real-time performance and high accuracy. Summary of the Invention
[0009] To address the aforementioned technical problems, this invention provides a fixed-time robust direct thrust control method and system for aero-engines. The technical means employed in this invention are as follows: A fixed-time robust direct thrust control system for an aero-engine includes: The linear model building module is used to collect data on the high-pressure rotor speed, low-pressure rotor speed, fuel flow and thrust of the aero-engine. Based on the equilibrium operating point of the preset cruise condition, the nonlinear model of the engine is linearized to build a linear state-space model containing sensor disturbances. The fixed-time thrust observer module is used to receive the state data output by the linear model. Through the combined effect of linear and nonlinear gains, it estimates the engine thrust, ensuring that the estimation error converges within a fixed time independent of the initial conditions, and constraining the impact of disturbances on the estimation error within a preset range. The direct thrust controller module receives the thrust estimate output by the observer, combines it with a preset reference thrust trajectory, and generates a fuel flow regulation signal through feedback control logic to achieve precise tracking of engine thrust. The gain solving module is used to construct optimization problems by pre-set stability criteria and constraints, and to solve the gain parameters of the observer and controller using numerical calculation tools. The electronic control unit (ECU) is used to realize data transmission and command execution of various modules, collect sensor signals in real time, drive the observer and controller to operate, and output adjustment signals to the engine actuator.
[0010] Furthermore, the gain of the fixed-time thrust observer module includes linear gain and nonlinear gain. The linear gain is designed to satisfy stability constraints by using a preset symmetric positive definite matrix, and the nonlinear gain is achieved by element-wise multiplication. Together, they make the upper bound of the convergence time of the thrust estimation error a preset fixed value.
[0011] Furthermore, the linear model construction module includes a multi-sensor data acquisition unit, wherein the sensors include a speed sensor, a fuel flow sensor, and a thrust indirect measurement sensor. After synchronous preprocessing, the acquired data is linearized through a first-order Taylor expansion to generate a linear state-space model containing a perturbation distribution matrix.
[0012] Furthermore, the direct thrust controller module adopts a direct feedback design of thrust estimation value, receives real-time thrust estimation data output by the observer, calculates the deviation with the preset reference thrust trajectory, and generates a fuel flow regulation signal through proportional adjustment and integral compensation. The reference thrust trajectory is constructed by a preset matrix of state variables, input variables and output variables.
[0013] Furthermore, the optimization problem constructed by the gain solving module includes symmetric matrix positive definite constraints, disturbance suppression performance constraints, and stability constraints. The linear gain matrix of the observer and the feedback gain matrix of the controller are obtained by numerical calculation tools. The nonlinear gain parameters of the observer are calculated by pre-given nonlinear gain matrices and constants. The above gains together ensure the asymptotic stability of the closed-loop system.
[0014] Furthermore, the linear model building module also includes a disturbance modeling unit, which uses a probabilistic noise model to characterize sensor measurement noise and optimizes the initial parameters of the model through maximum likelihood estimation to improve the model identification accuracy in noisy environments.
[0015] Furthermore, the direct thrust controller module also includes a closed-loop stabilization unit, which transforms the thrust tracking problem into a stabilization problem of the augmented system. The augmented state is constructed by a preset integral factor to ensure that the system state asymptotically converges to zero when there is no disturbance.
[0016] This invention also discloses a fixed-time robust direct thrust control method for aero-engines, employing the above-mentioned system and comprising the following steps: Step 1: Collect engine high-pressure rotor speed, low-pressure rotor speed, and fuel flow data through multiple sensors, simultaneously record steady-state operating point parameters, linearize the nonlinear model, and construct a linear state-space model with disturbances. Step 2: Start the fixed-time thrust observer, input the state data of the linear model, and output the thrust estimate through the preset linear gain and nonlinear gain co-operation to ensure that the estimation error converges to the allowable range within a fixed time. Step 3: The direct thrust controller receives the thrust estimate and the preset reference thrust trajectory, calculates the deviation, generates a fuel flow regulation signal through feedback control, and outputs it to the engine actuator; Step 4: Construct an optimization problem with stability constraints using the gain solver module, and use numerical calculation tools to solve for the gain parameters of the observer and controller, updating them periodically to adapt to changes in operating conditions; Step 5: The electronic control unit monitors the impact of disturbances in real time and adjusts the estimation accuracy through the disturbance suppression mechanism of the observer to ensure that the thrust tracking error meets the preset performance indicators.
[0017] Compared with the prior art, the present invention has the following advantages: 1. This invention relates to a novel direct thrust control method for aero-engines based on a fixed-time thrust observer, which can achieve fixed-time thrust tracking while ensuring L2 gain performance. The linear gain of the observer is designed using the Lyapunov function method, while the nonlinear gain is derived analytically. The rigorous derivation of the upper bound of the convergence time enhances the interpretability of the observer.
[0018] 2. This invention overcomes the inherent thrust-speed mismatch problem in the indirect thrust control architecture, thereby solving the thrust tracking problem of aero-engines. Sufficient conditions for the solvability of direct thrust control are established through Lyapunov analysis, strengthening the theoretical foundation.
[0019] 3. This invention uses an optimization-based method to efficiently solve for the controller gain, significantly reducing the computational resource requirements. Attached Figure Description
[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0021] Figure 1 This is a comparison chart of thrust estimation for this invention (with and without interference).
[0022] Figure 2This is a comparison chart of the rotational speed estimation of the present invention (with and without interference).
[0023] Figure 3 This is a comparison chart of the rotational speed estimation of the present invention (with and without interference).
[0024] Figure 4 This is a comparison chart of fuel estimation for the present invention (with / without interference).
[0025] Figure 5 This is a comparison chart of the thrust tracking performance of the present invention.
[0026] Figure 6 This is a comparison chart of fuel flow rates according to the present invention. Detailed Implementation
[0027] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.
[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0029] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.
[0030] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values of the components and steps described in these embodiments do not limit the scope of the invention. It should also be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale. Techniques, methods, and devices known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and devices should be considered part of the specification. In all examples shown and discussed herein, any specific values should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters in the following figures denote similar items; therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.
[0031] In the description of this invention, it should be understood that the orientation or positional relationship indicated by directional terms such as "front, back, up, down, left, right", "horizontal, vertical, horizontal" and "top, bottom" is generally based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing this invention and simplifying the description. Unless otherwise stated, these directional terms do not indicate or imply that the device or element referred to must have a specific orientation or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on the scope of protection of this invention. The directional terms "inner" and "outer" refer to the inner and outer contours relative to the outline of each component itself.
[0032] For ease of description, spatial relative terms such as "above," "over," "on the upper surface of," "above," etc., are used herein to describe the spatial positional relationship of a device or feature as shown in the figures to other devices or features. It should be understood that spatial relative terms are intended to encompass different orientations in use or operation besides the orientation of the device as described in the figures. For example, if the device in the figures is inverted, a device described as "above" or "above" other devices or structures would subsequently be positioned as "below" or "under" other devices or structures. Thus, the exemplary term "above" can include both "above" and "below." The device may also be positioned in other different ways (rotated 90 degrees or in other orientations), and the spatial relative descriptions used herein will be interpreted accordingly.
[0033] Furthermore, it should be noted that the use of terms such as "first" and "second" to define components is merely for the purpose of distinguishing the corresponding components. Unless otherwise stated, the above terms have no special meaning and therefore should not be construed as limiting the scope of protection of this invention.
[0034] like Figure 1As shown, this invention discloses a fixed-time robust direct thrust control system for an aero-engine, comprising: The linear model building module is used to collect data on the high-pressure rotor speed, low-pressure rotor speed, fuel flow and thrust of the aero-engine. Based on the equilibrium operating point of the preset cruise condition, the nonlinear model of the engine is linearized to build a linear state-space model containing sensor disturbances. The fixed-time thrust observer module is used to receive the state data output by the linear model. Through the combined effect of linear and nonlinear gains, it estimates the engine thrust, ensuring that the estimation error converges within a fixed time independent of the initial conditions, and constraining the impact of disturbances on the estimation error within a preset range. The direct thrust controller module receives the thrust estimate output by the observer, combines it with a preset reference thrust trajectory, and generates a fuel flow regulation signal through feedback control logic to achieve precise tracking of engine thrust. The gain solving module is used to construct optimization problems by pre-set stability criteria and constraints, and to solve the gain parameters of the observer and controller using numerical calculation tools. The electronic control unit (ECU) is used to realize data transmission and command execution of various modules, collect sensor signals in real time, drive the observer and controller to operate, and output adjustment signals to the engine actuator.
[0035] Furthermore, the gain of the fixed-time thrust observer module includes linear gain and nonlinear gain. The linear gain is designed to satisfy stability constraints by using a preset symmetric positive definite matrix, and the nonlinear gain is achieved by element-wise multiplication. Together, they make the upper bound of the convergence time of the thrust estimation error a preset fixed value.
[0036] Furthermore, the linear model construction module includes a multi-sensor data acquisition unit, wherein the sensors include a speed sensor, a fuel flow sensor, and a thrust indirect measurement sensor. After synchronous preprocessing, the acquired data is linearized through a first-order Taylor expansion to generate a linear state-space model containing a perturbation distribution matrix.
[0037] Furthermore, the direct thrust controller module adopts a direct feedback design of thrust estimation value, receives real-time thrust estimation data output by the observer, calculates the deviation with the preset reference thrust trajectory, and generates a fuel flow regulation signal through proportional adjustment and integral compensation. The reference thrust trajectory is constructed by a preset matrix of state variables, input variables and output variables.
[0038] Furthermore, the optimization problem constructed by the gain solving module includes symmetric matrix positive definite constraints, disturbance suppression performance constraints, and stability constraints. The linear gain matrix of the observer and the feedback gain matrix of the controller are obtained by numerical calculation tools. The nonlinear gain parameters of the observer are calculated by pre-given nonlinear gain matrices and constants. The above gains together ensure the asymptotic stability of the closed-loop system.
[0039] Furthermore, the linear model building module also includes a disturbance modeling unit, which uses a probabilistic noise model to characterize sensor measurement noise and optimizes the initial parameters of the model through maximum likelihood estimation to improve the model identification accuracy in noisy environments.
[0040] Furthermore, the direct thrust controller module also includes a closed-loop stabilization unit, which transforms the thrust tracking problem into a stabilization problem of the augmented system. The augmented state is constructed by a preset integral factor to ensure that the system state asymptotically converges to zero when there is no disturbance.
[0041] This invention also discloses a fixed-time robust direct thrust control method for aero-engines, employing the above-mentioned system and comprising the following steps: Step 1: Collect engine high-pressure rotor speed, low-pressure rotor speed, and fuel flow data through multiple sensors, simultaneously record steady-state operating point parameters, linearize the nonlinear model, and construct a linear state-space model with disturbances. Step 2: Start the fixed-time thrust observer, input the state data of the linear model, and output the thrust estimate through the preset linear gain and nonlinear gain co-operation to ensure that the estimation error converges to the allowable range within a fixed time. Step 3: The direct thrust controller receives the thrust estimate and the preset reference thrust trajectory, calculates the deviation, generates a fuel flow regulation signal through feedback control, and outputs it to the engine actuator; Step 4: Construct an optimization problem with stability constraints using the gain solver module, and use numerical calculation tools to solve for the gain parameters of the observer and controller, updating them periodically to adapt to changes in operating conditions; Step 5: The electronic control unit monitors the impact of disturbances in real time and adjusts the estimation accuracy through the disturbance suppression mechanism of the observer to ensure that the thrust tracking error meets the preset performance indicators.
[0042] Example 1 This embodiment provides a fixed-time robust direct thrust control method for a linear model of an aero-engine, including the following steps: S1: A model of the aero-engine was established, and the problems to be solved were proposed; In specific problem modeling, the following steps are included: S11: Establish a model of the aircraft engine: The nonlinear model of the aero-engine control system is as follows: (1) in, This refers to the high-voltage rotor speed. For low-pressure rotor speed, For fuel flow rate, For engine thrust, and Represents a nonlinear function with respect to the state. This represents a nonlinear function with respect to the input.
[0043] To construct a linear model under cruise conditions, an equilibrium operating point is first selected within a specified cruise range, and its steady-state data is recorded. Input quantities, system state quantities, and output quantities are collected synchronously to ensure data synchronization. Subsequently, based on this equilibrium point, a first-order Taylor expansion is performed on the nonlinear model (1) to linearize it, resulting in the following approximate linear state-space representation:
[0044] (2) in Represents partial derivatives, This represents the deviation between the actual value and the steady-state value. The engine control system acquires system information through multiple sensors, a process that inevitably introduces measurement noise and signal interference. To characterize the measurement process, this embodiment employs a probabilistic noise model and utilizes maximum likelihood estimation to optimize the initial parameter estimates, thereby improving identification accuracy under noisy conditions.
[0045] make , , The resulting linear aero-engine model, incorporating sensor noise, is as follows: (3) in, Indicates measurable output. This represents a perturbation. Assume there exist known positive constants. , making The system matrix is derived from the nonlinear model, and its specific form is as follows:
[0046] in, , , It is a perturbation distribution matrix with appropriate dimension.
[0047] S13: Control problems to be solved For the aero-engine control system (3), this embodiment aims to design a fixed-time thrust observer and a direct thrust controller, so that the system can still achieve accurate thrust estimation and tracking when disturbances exist. The observer needs to guarantee the fixed-time convergence of the estimation error, while providing... The controller should utilize the observer output to achieve accurate direct thrust tracking with ideal robustness, while maintaining horizontal disturbance suppression capability.
[0048] S2: Fusion A suppressed fixed-time thrust observer, this design reduces noise to estimation error. The gain is constrained within a preset range.
[0049] In a specific implementation example, the following steps are included. S21: Design a fixed time Thrust observer: The designed fixed-time thrust observer is as follows:
[0050]
[0051] (4) in , and They represent , and The observed values, For the linear gain of the observer, and For the nonlinear gain of the observer, , , , This represents element-wise multiplication. , is a given constant.
[0052] Designed fixed time The observer aims to achieve two main objectives: i) Under conditions of no external disturbance Under these conditions, the observer can accurately converge to the true thrust value within a fixed time, and the convergence time is independent of the initial conditions; ii) There is a non-zero external disturbance. At that time, the output estimation error Meet the following performance indicators: (5) in, To indicate The constant of gain.
[0053] To achieve the above control objectives, the following design process is adopted: Without considering nonlinear terms and In this case, the linear gain is first designed based on the Lyapunov function method. .
[0054] Define state estimation error and The error dynamic equation can be obtained as follows: (6) Theorem 1: For a thrust estimation system described by equation (6), if there exists a matrix... and normal numbers The following constraints must be met: (7) in It has been proven that the fixed-time gain design described below... The thrust observer (4) can converge in a fixed time independent of the initial conditions of the system. The internal convergence towards the true thrust value is achieved.
[0055] Prove to choose a Lyapunov function
[0056] in, It is a symmetric matrix.
[0057] function Regarding time The derivative is
[0058] in, .
[0059] It is not difficult to conclude
[0060] Next, we will process the nonlinear terms. .
[0061] make , , , , It is a diagonal matrix.
[0062]
[0063] because It is a positive definite symmetric matrix, from which we can derive...
[0064] There are also
[0065] Therefore, it can be concluded that
[0066] Next, we will process the nonlinear terms.
[0067] It is not difficult to conclude
[0068]
[0069] Right now
[0070] Therefore, it is easy to obtain
[0071] in
[0072] In addition, it can be deduced that
[0073]
[0074] in
[0075] From the conditions of Theorem 1, we know that Substituting the above inequality into the equation, we get:
[0076] in .
[0077] For both sides of the equation in the time interval Integrating the above, we get:
[0078] Further derivation yields:
[0079] Taking advantage of the boundedness of the disturbance and its fixed-time convergence property, the following inequality holds when the external disturbance is zero:
[0080] Therefore, we can conclude that:
[0081] in .in Represents any point in time. This is the upper bound of the time for fixed-time convergence.
[0082] The gain property has been proven. Now we prove the fixed-time convergence. When At that time, there were:
[0083] review It can be deduced that:
[0084] From this, we can derive the upper bound of fixed-time convergence. satisfy:
[0085] Therefore, the state estimation error Available at a fixed time It converges to zero, and Compared with the initial state Irrelevant.
[0086] In summary, the error system (6) has a fixed-time convergence characteristic, and its convergence time does not exceed [a certain value]. .
[0087] It is important to note that the observer gain structure contains both linear terms and key nonlinear terms: 1) Linear gain: .
[0088] 2) Nonlinear gain: and .
[0089] The upper bound for fixed-time convergence satisfies the following condition: (8) in , .
[0090] S22: Direct Thrust Control Design S221: Description of Direct Thrust Control Model
[0091] (9) in, express The estimated value, express The estimated value, For a matrix of appropriate dimension, Indicates a disturbance.
[0092] The reference thrust trajectory is defined as follows:
[0093] (10) in, , , These represent the reference system state variables, input variables, and output variables, respectively. , and Let be a matrix of appropriate dimension. , ,in Indicates the initial time. is the integrating factor.
[0094] S221: Direct Thrust Controller Design To achieve thrust tracking control directly using thrust, the following direct thrust controller is designed: (11) in , and It is the gain of the controller to be designed.
[0095] Substituting the controller into system (9), the closed-loop system model can be obtained as follows:
[0096] (12) based on From the expression, we can obtain (13) make , Combining (10), (12), and (13), we can conclude that
[0097] (14) in
[0098] Therefore, the original control problem has now been transformed into a stabilization problem for the augmented system (13). The control objective has two levels: to achieve accurate thrust tracking by using direct feedback control that estimates the thrust; specifically: 1) When there is no disturbance, the system state asymptotically converges to zero; 2) When disturbances exist, the thrust estimation error satisfies the following: Gain performance metrics: (15) in express Gain.
[0099] S3: Controller Design and Controller Gain Determination S31: Design the controller and determine the controller gain using a convex optimization framework.
[0100] Theorem 2 For system (14), if there exists a symmetric matrix , matrix , and and normal numbers The following constraints must be met: (16) in
[0101] The proposed controller (11) enables the closed-loop system (12) to be globally asymptotically stable without disturbance, and satisfies the following when there is a disturbance: Gain performance metrics.
[0102] Proof of the choice of Lyapunov function
[0103] in , , and It is the matrix to be designed.
[0104] according to and The specific form can be derived as follows:
[0105] For the Lyapunov function (20) with respect to time Taking the derivative, we get:
[0106] According to (16), when At that time, it can be concluded that Therefore, it can be determined that the system has asymptotic stability.
[0107] At this point, considering both the system output and the disturbance signal, we can obtain:
[0108] Based on equation (16), when At that time, it can be concluded that
[0109] Integrating both sides of the above equation, we can see that:
[0110] According to the definition of the Lyapunov function, when time... As we approach infinity, we obtain:
[0111] Therefore, it can be determined that... The gain performance indicators meet the requirements, and the proof is complete.
[0112] S32: Controller Gain Calculation Since the constraint condition (16) in Theorem 2 has non-convex characteristics, it is difficult to directly solve the controller gain. Now, the following theorem is used to transform it into a solvable form, so as to effectively calculate the gain matrix.
[0113] Theorem 3 For system (14), if there exists a symmetric matrix matrix , , and normal numbers It satisfies the following constraints: (17) in
[0114] The proposed controller (11) can be solved by explicit gain. , and Simultaneously solve the output tracking problem of the closed-loop system (12). Control issues, ensuring the system's global asymptotic stability and preset parameters. Gain performance.
[0115] The proof is that by applying Schul's complement lemma to inequality (16), we obtain:
[0116] Each diagonal matrix is defined as follows:
[0117] To decouple the Lyapunov matrix from the controller gain and obtain the linear form of the decision variables, left and right multiplications are performed on (21). And introduce variable substitution:
[0118] Under this contractual transformation, the inequality retains its sign and is transformed into the form (17) under the new variables, which is convex and numerically solvable. Theorem 3 is thus proved.
[0119] To verify the effectiveness of the control strategy proposed in this example, a hardware-in-the-loop experiment was conducted using a certain type of aero-engine as an example. Based on the engine model and the system identification method described in S1, the engine model was linearized at the steady-state operating point, resulting in the following linear model parameters for the aero-engine:
[0120] Subsequently, the measurement output matrix and the disturbance matrix are defined as follows:
[0121] By solving equation (7) in Theorem 1, the linear observer gain matrix is obtained. With Lyapunov function matrix as follows:
[0122] Given a nonlinear gain matrix , and constants and The nonlinear gain and the fixed-time convergence upper bound can be calculated as follows:
[0123] To verify the proposed fixed time The effectiveness and superiority of the thrust observer are compared with thrust estimation methods based on Kalman filtering (without considering measurement errors) and methods based on... A comparison is made between the thrust estimation method using Kalman filtering (considering measurement errors). Experimental results are as follows: Figures 1-6 As shown.
[0124] Figure 1 A comparison of thrust estimation results is presented. Under undisturbed conditions, the proposed fixed-time... Thrust estimation methods can reduce the estimation error to within Pre-convergence, thrust estimation methods based on Kalman filtering only converge in... The proposed method converges later. When measurement errors exist, the proposed method exhibits smaller error fluctuations, and its estimates are less affected by disturbances.
[0125] Figure 2 and Figure 3 The comparison results of the estimated rotor speeds for high / low pressure under undisturbed and disturbed conditions are presented respectively. For example... Figure 2 As shown, under undisturbed conditions, the proposed fixed time... Thrust estimation methods can achieve faster convergence of rotational speed estimation. Figure 3 The results show that, in the presence of disturbances, the proposed method exhibits smaller fluctuations in speed estimation and is less sensitive to disturbances (i.e., it has better anti-interference capabilities).
[0126] Fuel flow estimation results are as follows Figure 4 As shown, the proposed method exhibits smaller estimation errors and lower fuel flow fluctuations, achieving superior fuel flow control performance.
[0127] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A fixed-time robust direct thrust control method and system for an aero-engine, characterized in that, include: The linear model building module is used to collect data on the high-pressure rotor speed, low-pressure rotor speed, fuel flow and thrust of the aero-engine. Based on the equilibrium operating point of the preset cruise condition, the nonlinear model of the engine is linearized to build a linear state-space model containing sensor disturbances. The fixed-time thrust observer module is used to receive the state data output by the linear model. Through the combined effect of linear and nonlinear gains, it estimates the engine thrust, ensuring that the estimation error converges within a fixed time independent of the initial conditions, and constraining the impact of disturbances on the estimation error within a preset range. The direct thrust controller module receives the thrust estimate output by the observer, combines it with a preset reference thrust trajectory, and generates a fuel flow regulation signal through feedback control logic to achieve precise tracking of engine thrust. The gain solving module is used to construct optimization problems by pre-set stability criteria and constraints, and to solve the gain parameters of the observer and controller using numerical calculation tools. The electronic control unit (ECU) is used to realize data transmission and command execution of various modules, collect sensor signals in real time, drive the observer and controller to operate, and output adjustment signals to the engine actuator.
2. The system according to claim 1, characterized in that, The gain of the fixed-time thrust observer module includes linear gain and nonlinear gain. The linear gain is designed to meet stability constraints by using a preset symmetric positive definite matrix, and the nonlinear gain is achieved by element-wise multiplication. Together, they make the upper bound of the convergence time of the thrust estimation error a preset fixed value.
3. The system according to claim 1, characterized in that, The linear model construction module includes a multi-sensor data acquisition unit, which includes a speed sensor, a fuel flow sensor, and a thrust indirect measurement sensor. After synchronous preprocessing, the acquired data is linearized through a first-order Taylor expansion to generate a linear state-space model containing a perturbation distribution matrix.
4. The system according to claim 1, characterized in that, The direct thrust controller module adopts a direct feedback design of thrust estimation value, receives real-time thrust estimation data output by the observer, calculates the deviation with the preset reference thrust trajectory, and generates a fuel flow regulation signal through proportional adjustment and integral compensation. The reference thrust trajectory is constructed by a preset matrix of state variables, input variables and output variables.
5. The system according to claim 1, characterized in that, The optimization problem constructed by the gain solving module includes symmetric matrix positive definite constraints, stability constraints, and disturbance suppression performance constraints. The linear gain matrix of the observer and the feedback gain matrix of the controller are obtained by numerical calculation tools. The nonlinear gain parameters of the observer are calculated by pre-given nonlinear gain matrices and constants. The above gains together ensure the asymptotic stability of the closed-loop system.
6. The system according to claim 1, characterized in that, The linear model building module also includes a disturbance modeling unit, which uses a probabilistic noise model to characterize sensor measurement noise and optimizes the initial parameters of the model through maximum likelihood estimation to improve the model identification accuracy in noisy environments.
7. The system according to claim 1, characterized in that, The direct thrust controller module also includes a closed-loop stabilization unit, which transforms the thrust tracking problem into a stabilization problem of the augmented system. The augmented state is constructed by a preset integral factor to ensure that the system state asymptotically converges to zero when there is no disturbance.
8. A fixed-time robust direct thrust control method for an aero-engine as described in any one of claims 1-7, characterized in that, Includes the following steps: Step 1: Collect engine high-pressure rotor speed, low-pressure rotor speed, and fuel flow data through multiple sensors, simultaneously record steady-state operating point parameters, linearize the nonlinear model, and construct a linear state-space model with disturbances. Step 2: Start the fixed-time thrust observer, input the state data of the linear model, and output the thrust estimate through the preset linear gain and nonlinear gain co-operation to ensure that the estimation error converges to the allowable range within a fixed time. Step 3: The direct thrust controller receives the thrust estimate and the preset reference thrust trajectory, calculates the deviation, generates a fuel flow regulation signal through feedback control, and outputs it to the engine actuator; Step 4: Construct an optimization problem with stability constraints using the gain solver module, and use numerical calculation tools to solve for the gain parameters of the observer and controller, updating them periodically to adapt to changes in operating conditions; Step 5: The electronic control unit monitors the impact of disturbances in real time and adjusts the estimation accuracy through the disturbance suppression mechanism of the observer to ensure that the thrust tracking error meets the preset performance indicators.
