Vertical Takeoff and Vertical Landing (VTVL) reusable rockets are pivotal to sustainable space exploration, yet their recovery process presents critical challenges in rapid-response, high-precision attitude control. To address these challenges, this paper proposes a predefined-time attitude control scheme that integrates adaptive state constraints with explicit input-saturation compensation. First, Predefined-Time Extended State Observers (PTESOs) are designed for rapid and accurate disturbance estimation. Subsequently, an Adaptive Time-varying Barrier Lyapunov Function (ATBLF) is proposed to address abrupt state changes beyond preset constraints, offering an enhanced alternative to conventional state-constraint approaches. Building on these components, a Predefined-Time Control (PTC) approach that incorporates a Predefined-Time Anti-Saturation Compensator (PTASC) is presented, which guarantees predefined-time convergence while ensuring fin deflection angles remain within physical amplitude limits. Finally, comparative numerical simulations demonstrate the superiority of the proposed method in terms of transient response, tracking accuracy, and chattering elimination. (c) 2026 COSPAR. Published by Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.

Aiming at the typical failure problems of normally open and normally closed attitude control nozzles of air and space vehicles, the CNN algorithm is used to extract high-dimensional features from the time series data, and the LSTM model is used to learn the temporal features of the time series data, so as to establish a CNN-LSTM nozzle failure prediction model, and carry out the prediction analysis. The results show that compared with the two benchmark models of CNN and LSTM, the CNN-LSTM prediction model has smaller mean absolute error rate and mean square error rate, and the prediction effect is better.

Model predictive control, as a model-based algorithm, requires accurate motor parameters for its high-performance control, but the parameter mismatch caused by the insufficiency of a priori knowledge as well as the nonlinear changes of parameters with the operating environment and conditions will affect the accuracy of the predictive model and greatly degrade the control performance of the motor system. In response to the above problems, we take the dual three-phase permanent magnet synchronous motor as the research object, and proposed an adaptive disturbance observer-based model predictive control method. The error caused by parameter mismatch is estimated and compensated by an online adjustable neural network observer, which is optimized by a neural network weight tuning law designed using Lyapunov theory, ensuring the convergence of the weights and the stability of the observer. Experimental results demonstrate the effectiveness of the proposed method.

Data mining is one of the hot research topics in the field of databases in recent years. Association analysis is one of the key techniques in data mining, which mainly involves identifying all frequent itemsets in the database, and then generating association rules from the frequent itemsets. This article takes aircraft as the research object, intelligently processes a large amount of experimental data of aircraft control systems, and studies association rule mining methods based on discrete and mixed data, as well as association rule mining methods based on continuous data. For the case of discrete and mixed data, research is conducted based on the Apriori algorithm to obtain corresponding association rules. For continuous data types, particle swarm optimization (PSO) algorithm is used for processing to obtain corresponding association rules. Then, association rules are mined from a large amount of data to abstract design experiences that are conducive to judgment and comparison, improving the efficiency and speed of control system design. Finally, the effectiveness of the algorithm is verified through simulation experiments and analysis.

High reliability and intelligence mark the cutting-edge development trend of high-speed vehicle control technology. This research focuses on the progress of high-speed vehicle control technology based on reinforcement learning, and sorts out its application in control parameter tuning, compensation control, and end-to-end control. Focusing on the development framework of intelligent control safety and efficiency, this paper deeply analyzes the challenges of safety, training efficiency, and interpretability faced by high-speed vehicle in reinforcement learning control, and proposes feasible paths and development directions for future research.

High-speed Morphing Vehicles (HMVs) embody advanced biomimetic technology but pose significant control challenges due to their fast-varying dynamics, inherent uncertainties, and stringent efficiency requirements. These issues necessitate an advanced attitude control system capable of rapid response, effective disturbance rejection, and green-efficient operation. This paper proposes a dual-loop identifier-actor-critic structure for the attitude control system of HMVs to satisfy these demands. Within this structure, a fixed-time Neural Network (NN)-based identifier provides feedforward compensation to accurately estimate and mitigate system uncertainties in real time. Based on this, a novel fixed-time optimal backstepping control algorithm is developed using actor and critic NNs, which solves the Hamilton-Jacobi-Bellman equation online to achieve optimal performance. All these NNs update their weights online via adaptive laws to enable rapid convergence. Detailed comparative simulations validate the superior effectiveness of the proposed method, confirming the improvements in both tracking performance and energy efficiency for HMV attitude control.

This paper introduces a distributed cooperative guidance law that uses a parallel state observer (PSO) in guidance systems, allowing all missiles to intercept a maneuvering target at a specific terminal angle. The use of a parallel state observer, which consists of multiple differentiators connected simultaneously, can improve the fault tolerance of the guidance system. The leader-follower topology is utilized for transmitting information among multiple missiles. The cooperative control algorithm integrates PSO with a nonsingular feedback linearization strategy. Simulation examples are provided to illustrate the effectiveness of the proposed guidance method.

In this paper, a data-based BP neural network offline parameter identification method is proposed with a 500kg-class rotorcraft as the research object. Firstly, multiple manned flight tests are conducted to obtain the flight test data for offline training; secondly, the longitudinal and lateral identification models of aerodynamic data are established by offline training using BP neural network algorithm with the pilot's action of the operating stick as the input and the flight state quantities of the actual flights as the output; lastly, the nonlinear mathematical simulation is conducted to validate the effectiveness of the identification algorithm by comparing the flight test and the flight state quantities solved by the identification model, and the CR bounds are used as the analytical method to verify the validity of the identification algorithm. Finally, nonlinear mathematical simulation is carried out to verify the effectiveness of the recognition algorithm by comparing the flight test and the state quantities solved by the recognition model, and the recognition accuracy is evaluated by using the CR community as an analytical tool.

In order to meet the needs of attitude anomaly detection of aerocraft, this paper designs a data-driven approach for aerocraft attitude information anomaly detection, which realizes the detection and analysis of data feature information based on the time domain and frequency domain analysis of data, in which the small jitter detection part uses the frequency domain analysis of attitude data based on short-time Fourier transform, and the segmentation setting of power threshold improves the accuracy and reliability of jitter detection. The attitude instability detection part selects the mode and sets the threshold detection range according to the data characteristics, which can detect the obvious attitude instability phenomenon of the data. Finally, a semi-physical simulation system for space vehicles was built for experimental verification, which proved the effectiveness of the algorithm.

Morphing aircraft exhibit agile maneuverability within an extensive flight envelope, which imposes heightened requirements on the online adaptation capabilities of control laws and necessitates enhanced dynamic flight control performance. The coefficient freezing method frequently used in engineering applications entails individualized design procedures for each distinct operating point, a process that is both intricate and heavily reliant on human expertise. Moreover, once deployed, conventional flight control laws generally exhibit static performance with limited potential for online optimization and adjustment. This work presents an offline strategy gradient deep reinforcement learning approach to devise a self-learning architecture for the parameters of PI state feedback control laws. This architecture is trained automatically by traversing across multiple operational points of the morphing aircraft under control, thereby yielding a set of control laws adaptable to various operational conditions. Concurrently, the action network from this framework is extracted and implemented as an online learning module, enabling real-time adjustments to control law parameters during flight based on the observed control performance, thus augmenting the aircraft’s inherent online adaptive capabilities. The validity and effectiveness of the proposed methodology are substantiated through a case study focusing on tracking control in the pitch channel of a representative morphing aircraft.

This paper investigates the finite-time position trajectory tracking control problem of quadrotor unmanned aerial vehicles (UAVs). Different from the standard inner–outer-loop control scheme, the proposed finite-time controller is constructed with an order-supplementary mechanism. Concretely, some virtual extended states with second-order dynamics are utilized in the controller design of the UAV’s position-loop subsystem, to replace the original feedback part of tracking errors. Then, the adding a power integrator technique is used in the establishment of the virtual state dynamics, such that the position loop of quadrotor UAVs achieves the trajectory tracking tasks in finite time. Meanwhile, the attitude command references are directly formulated from the virtual extended states. Moreover, to deal with disturbances or unknown velocities, some finite-time observers are further combined in the proposed approach to obtain the corresponding estimates of disturbances and velocities. Compared with the existing results, the proposed order-supplementary finite-time trajectory tracking approach can remove the use of filters in the attitude command resolution and realize strict finite-time convergence. The thrust control input for the position-loop subsystem can be adjusted more flexibly by setting the initial values of the introduced virtual states. In addition, the finite-time velocity observer provided in this paper takes aerodynamic damping into account and has more accurate estimation in practice. Some simulations are given to validate the effectiveness of the proposed approach.

The fault tolerant control (FTC) of hypersonic flight vehicle (HFV) with actuator fault is proposed in this paper. The fault model considered in this paper is a general model which HFV may encounter in practice. For designing the FTC, the nonlinear model of HFV is represented by T-S fuzzy models, and policy iteration (PI) strategy is utilized to solve the design problem of T-S controller for the built T-S model of HFV without actuator fault. Then, based on the normal T-S controller, an adaptive fuzzy FTC controller is proposed, in which the feedback gain matrices can improve themselves according to the special fault. The stability of the proposed adaptive fuzzy FTC is proved by Lyapunov theory, and an integral reinforcement learning (IRL)–based solving algorithm is proposed. Simulations on three different kinds of actuator faults are proposed, and the simulation results show that, under three different faults, the designed adaptive FTC can ensure the reliable flight of HFV.

In order to enhance the matching relationship between guidance subsystem and control subsystem of the thrust-vector-controlled aircraft, a kind of sufficient modeling and adaptive robust design facing to integrated guidance and control (IGC) is proposed. With respect to the researched aircraft, a linear state-dependent mathematical model facing to IGC design is first established. Based on the as-built model, a new kind of adaptive robust IGC law is proposed and it is composed by an adaptive optimal IGC law and a robustness-improved IGC law. In order to guarantee the global stability of time-varying control system, the adaptive optimal IGC law is designed by solving Riccati matrix equation on line and using matrix Sign function method. Furthermore, in order to enhance the robust ability against the unmatched system uncertainties, the robustness-improved IGC law is designed by using dynamic surface control approach and extended state observing strategy. Simulation results present that, the proposed IGC scheme presents more performance advantages compared with traditional IGC schemes, including the improvement of guidance precision and attitude stabilization. Furthermore, in the conditions of 256 simulation combinations, the minimum relative distances between aircraft position and target position are distributed from 0.103m to 5.333m, and the average value is 2.208m, which means the proposed IGC scheme possesses strong robustness against different and time-varying model uncertainties.

The near-Earth space holds significant strategic value, and X-ray pulsar navigation in this region can diversify navigation methods, enhancing the safety and autonomy of spacecraft. Due to the challenges and high costs associated with obtaining actual measurements of pulsars in near-Earth space, the simulation technology for X-ray pulsar signals in this region can provide input for demonstrating the feasibility of technical solutions and experiments related to the application of X-ray pulsar navigation in near-Earth space. Therefore, this paper proposes a method for simulating X-ray signals in near-Earth space. Firstly, a scale transforming method is employed to generate a photon arrival time sequence at the spacecraft that includes energy information. Subsequently, the atmospheric transmittance model is utilized for photon selection, obtaining a sequence of photons that have passed through the atmosphere after absorption. Finally, experimental validation of the proposed algorithm is conducted using actual Crab pulsar data measured by the HXMT. The similarity between the simulated data and the measured data is evaluated in terms of profiles, phases, and energy spectra. Simulation experiments demonstrate that the time delay estimation error caused by the simulation algorithm is less than 17.17 us. When the observation tangent point altitude is in the range of 160 km to 180 km, the Pearson correlation coefficient of the profiles between simulated and measured data is 73.74%. The Pearson residuals of the energy spectra are evenly distributed between -0.2 and 0.2, indicating a good level of similarity between the two.

This article proposes an online universal self-learning control (USLC) algorithm based on a physical performance policy-optimization neural network, which aims to solve the problem of universal self-learning optimal control laws for nonlinear systems with various uncertain dynamics. As a key system characterization, this algorithm predicts the discrepancy between the optimal and current control laws by evaluating overall performance in each iterative learning cycle, leveraging an offline-trained universal policy network. This approach is universal, as it does not rely on an exact system model and can adaptively control performance preferences across various tasks by customizing the physical performance cost weights. Using the established control law-performance surface and contraction Lyapunov function, the necessary assumptions and proofs for the stable convergence of the system within a three-dimensional manifold space are provided. To demonstrate the universality of USLC, simulation experiments are conducted on two different systems: a low-order circuit system and a high-order variable-span aircraft attitude control system. The stable control achieved under varying initial values and boundary conditions in each system illustrates the effectiveness of the proposed method. Finally, the limitations of this study are discussed.

针对航天器姿态跟踪问题,建立了基于误差四元数的姿态控制系统二阶数学模型.基于现代控制理论,通过设计四元数状态反馈控制器对姿态控制系统的极点进行配置,实现航天器姿态的稳定跟踪.仿真结果表明,设计的状态反馈控制器能够实现微型航天器的高精度姿态跟踪控制.

This paper applies the adaptive sliding mode controller (ASMC) with Radial Basis Function (RBF) approximation for hypersonic unmanned flight vehicle (HFV) for attitude regulating and tracking. Combining the RBF capability of approximating nonlinear function online and the remarkable robustness of sliding mode variable structure, the proposed controller can be implemented on HFV with Lyapunov asymptotically stability even the plant is exerted with disturbance. First, the simplified longitudinal hypersonic flight vehicle and the dynamics of roll at low-frequency are given as model. Second, the proposed controller is designed with specific details, and the stability analysis is given as well. Third, two cases of numerical simulation of both roll angle regulating and angle of attack tracking in longitudinal plane prove the efficiency of this controller, and the input of actuator does not exceed the deflection angle limitation in both cases.

The cooperative control technology for swarming vehicles is analyzed and prospected systematically. The advanced swarming vehicles and cooperative control technology are summarized and compared, and five major scientific problems in this field are put forward. On this basis, the autonomous control system architecture for the vehicle swarm and kinds of key technologies, such as pre-launch planning, online situational awareness, cooperative guidance and control, are proposed. A brief overview of each technology is given, explaining the relationship between them and their roles in the cooperative control system. Finally, the future development of this field is prospected from three aspects of theoretical research, technological breakthrough and engineering practice.

Aiming at the problem of low accuracy of traditional fitting method for the wind field of launch vehicles, a high-precision fitting approach based on least parameter neural network is proposed. In this method, the flight altitude of launch vehicles is taken as the network input, and the wind field velocity and orientation are taken as the network outputs. The wind field fitting is completed with the minimum number of network layers and the number of neurons, and the lower bound formula of the number of hidden layer nodes is given. Compared with the traditional least-square polynomial fitting and its multi-segment style, the least-parameter network fitting can improve the precision with a unified framework. A large number of simulations results fully demonstrate the effectiveness, conciseness and robustness of the proposed method.

Brain-computer Interface (BCI) system based on motor imagery (MI) heavily relies on electroencephalography (EEG) recognition with high accuracy. However, modeling and classification of MI EEG signals remains a challenging task due to the non-linear and non-stationary characteristics of the signals. In this paper, a new time-varying modeling framework combining multiwavelet basis functions and regularized orthogonal forward regression (ROFR) algorithm is proposed for the characterization and classification of MI EEG signals. Firstly, the time-varying coefficients of the time-varying autoregressive (TVAR) model are precisely approximated with the multiwavelet basis functions. Then a powerful ROFR algorithm is employed to dramatically alleviate the redundant model structure and accurately recover the relevant time-varying model parameters to obtain high resolution power spectral density (PSD) features. Finally, the features are sent to different classifiers for the classification task. To effectively improve the accuracy of classification, a principal component analysis (PCA) algorithm is utilized to determine the best feature subset and Bayesian optimization algorithm is performed to obtain the optimal parameters of the classifier. The proposed method achieves satisfactory classification accuracy on the public BCI Competition II Dataset III, which proves that this method potentially improves the recognition accuracy of MI EEG signals, and has great significance for the construction of BCI system based on MI.