As multifunction radar performance continues to improve, there is an increasing number of instances of overload when a multifunction radar performs multiple tasks. Consequently, the radar task scheduling problem has become increasingly significant. In this paper, a multifunction radar scheduling model and a target value degree assessment system are developed based on intuitionistic fuzzy reasoning. Eight generic state features are extracted from the radar system to represent its status at each scheduling point, and a deep reinforcement learning based multifunction radar task resource scheduling algorithm is proposed. The efficacy and advantages of the algorithm are validated in five distinct task overload scenarios. The outcomes demonstrate that the algorithm is capable of optimal scheduling in varying sizes of task overload environments and reduces the total delay rate of task scheduling.
This paper addresses the problem of designing distributed cooperative guidance laws for multiple missiles in threedimensional (3D) space, considering the constraint of simultaneous impact and fully intermittent communication. The problem is structured using a 3D nonlinear mathematical model along with a topology network for interaction. By employing the relative distance and velocity leading angle as coordination variables, a novel finite-time cooperative guidance law is developed through an event-triggered strategy, which eliminates the need for timeto-go estimation. The law's convergence is demonstrated through finite-time Lyapunov stability theory. Additionally, the guidance approach also eliminates Zeno behavior, which refers to infinite communication instances within a finite time interval, ensuring fully intermittent communication among the missiles. To prevent singularity caused by non-zero velocity leading angles in cooperative guidance, the proportional navigation guidance (PNG) law is applied when coordination variables reach consensus and missiles are nearing the target, which can guarantee that all missiles strike the target simultaneously. Simulation rection within a finite duration while maintaining fully intermittent communication throughout the process.
The temporal-spatial focusing characteristic of time reversal technology has unique advantages in spatial power coherent synthesis. This paper studies the adaptive spatial power coherent synthesis technology based on time-reversal, discusses and analyses the influence of random phase errors and sampling time width on the signal synthesis effect at the target radar, and combines with simulation experiments to provide the principles for controlling random phase errors and selecting sampling time width, thereby providing a theoretical basis for the engineering application of time-reversal coherent synthesis technology.
In order to solve nonlinear problem caused by different coordinates in measurement and target state models during radar target tracking, a converted measurement Kalman filtering algorithm based on Taylor re-expansion with the predicted information (TREP-CMKF) is proposed in this paper. Based on second-order Taylor expansion, the first and second moments associated with the measurement conversion error are obtained, which eliminates the effect of measurement noise on error covariance by using predicted information. At the same time, the first-order expansion error in second order Taylor expansion is considered based on the predicted information to further reduce the conversion error and improve the consistency. In addition, the conversion and estimation error ellipses in Cartesian and polar coordinates are analyzed, and the hybrid Kalman filtering (HybridKF) algorithm is proposed, which combines the advantages of converted measurement method and square-root cubature Kalman filtering. Simulation results show that the proposed TREP-CMKF algorithm has good consistency and credibility in both measurement conversion and filter estimation, and the HybridKF algorithm has the highest tracking accuracy in all dimensions compared with existing converted measurement methods and nonlinear filter algorithms.
This study addresses electromagnetic scattering modeling of irregularly shaped particle clouds formed from air target disintegration at high altitude and proposes an integrated approach that couples dynamics with electromagnetic computation. The dynamic distribution of particles under the influence of both gravity and air resistance is simulated by means of the explosive motion equation. The geometric shape of the particle clouds is constructed using a three-dimensional convex hull algorithm; the particle clouds are partitioned based on the Kmeans clustering technique to calculate the local density. A generalized Rayleigh-Gans (GRG) approximation is used to obtain the scattering characteristics for single particles. To account for multiple scattering within the cloud, the Monte Carlo technique is employed under the framework of vector radiative transfer theory, which involves statistically sampling key photon parameters (entry points, free paths, and scattering angles) to estimate the full-angle scattering echoes. Moreover, simulation indicates that the radar cross-section of particle clouds is positively correlated with factors such as frequency and particle size. This method is effective for analyzing the electromagnetic scattering characteristics of particle clouds with complex shapes.
