To address the limitations of existing sliding mode-based integrated guidance and control (IGC) schemes, such as chattering, input saturation, and insufficient robustness, this paper proposes a three-dimensional IGC design method incorporating both terminal angle and attitude angle constraints. First, a control-oriented six-degrees-of-freedom model is established based on three-dimensional relative motion and vehicle dynamics, and the control objectives for maneuvering target interception under multiple constraints are clarified. Subsequently, a finite-time terminal sliding mode guidance law based on time-to-go (TGO) is integrated with dynamic surface control to construct the IGC framework. In this design, command filters are introduced to overcome the “explosion of complexity”, while amplitude saturation functions are employed to constrain system states and control inputs. Meanwhile, a generalized super-twisting extended state observer (GSTESO) is incorporated to estimate and compensate for lumped uncertainties in the system. Finally, by combining Lyapunov stability theory with an integral barrier Lyapunov (IBL) function, it is proven that the closed-loop system is uniformly ultimately bounded and satisfies the terminal angle constraints. Comparative simulations under multiple disturbance scenarios demonstrate that the proposed method meets the accuracy requirements in terms of miss distance and LOS angle error. Moreover, it alleviates high-frequency chattering and prevents control-input saturation, showing improved robustness and disturbance rejection capability compared with the baseline methods. Therefore, the proposed approach provides a valuable reference for engineering applications of three-dimensional IGC in maneuvering target interception.

For a class of second-order nonlinear systems subject to matched disturbances, this paper investigates preset-time establishment and subsequent high-precision maintenance, and proposes a time-varying sliding mode control method based on prescribed error evolution. First, a quintic prescribed error-evolution function satisfying smooth endpoint constraints is constructed to explicitly shape the error establishment process, thereby avoiding excessive transient control demand caused by directly injecting a large initial error into the closed loop. Then, a time-varying sliding variable is constructed around the error-evolution deviation, and a continuous control law is designed by combining a bounded nonlinear mapping with a distance-dependent scheduling gain, achieving coordination between preset-time establishment and input smoothness. The theoretical analysis proves that, under nominal conditions, the closed-loop system can complete the establishment exactly at the preset instant; in the presence of bounded matched disturbances, the system satisfies an explicit small-neighborhood estimate at the preset instant and maintains high precision thereafter. The simulation results show that the proposed method achieves timing-consistent state establishment under different preset establishment scenarios, and significantly reduces the control peak and total input variation, thereby effectively alleviating the risk of actuator saturation.

Multi-missile interception of maneuvering targets requires accurate prediction and impact-time coordination under fixed missile speed. This study couples LSTM-IMM tracking with active curvature planning. LSTM generates a time-varying IMM transition matrix from measurement features, and the guidance law converts fixed-speed coordination into curvature and normal-acceleration allocation. Comparisons, sensitivity tests, Monte Carlo simulations, and five engagement scenarios verify tracking, mode recognition, real-time feasibility, and interception without speed regulation.

This paper proposes a barrier-modulated probabilistic super-twisting guidance (BMPSTG) for robust three-dimensional maneuvering target interception under measurement uncertainty and packet loss. A stochastic barrier-adaptive mechanism inspired by tunneling dynamics is introduced to reshape the sliding variable evolution, enabling accelerated convergence and reduced chattering compared with conventional super-twisting schemes. To enhance disturbance reconstruction accuracy, a composite observer integrating a finite-time disturbance observer (FTDO) and a high-order extended state observer (HESO) is developed. The FTDO ensures fast transient estimation, while the HESO improves steady-state precision and robustness to noise. Simulation results demonstrate that the proposed FTDO–HESO structure outperforms both standalone FTDO and conventional FTDO–ESO configurations in terms of estimation accuracy and guidance performance. A probability-coupled gain adaptation strategy further adjusts control gains according to engagement states, improving robustness under aggressive target maneuvers. Monte Carlo simulations verify that the proposed method achieves high interception accuracy with stable convergence and reduced control effort in complex engagement scenarios.