To solve the target tracking problem with little a priori information about the target dynamics, our series of studies, including this paper as the third part, propose a continuous-time trajectory estimation approach (dubbed targeting track) based on the stochastic process (SP) theory and a deterministic-stochastic decomposition framework. Specifically, we decompose the learning of the trajectory SP into two sequential stages: the first fits the deterministic trend of the trajectory using a curve function of time, while the second estimates the residual stochastic component through learning either a Gaussian process (GP) or Student’s-t process (StP). The former has been addressed in the companion paper and the latter is the focus of this paper. This leads to a data-driven tracking approach that produces the continuous-time trajectory with minimal prior knowledge of the target dynamics. Notably, our approach models the temporal correlations of the state sequence and of measurement noise using separate GP or StP. It does not only take advantage of the smooth trend of the target but also makes use of the long-term temporal correlation of both the data and the model fitting error. Although the GP admits an exact closed-form expression for the linear system, approximations have to be adopted for StP modeling. Simulations in four maneuvering target tracking scenarios have demonstrated its effectiveness and superiority in comparison with existing approaches.

Target localization plays a key role in unmanned aerial vehicle (UAV) reconnaissance systems. Map-based target localization methods offer the advantage of being unaffected by installation and measurement errors. However, images captured by UAVs in complex ground environments often contain high-density feature points. Moreover, variations in illumination and cross-view between the target and reference images frequently introduce significant noise. These factors make it challenging for traditional map-based target localization methods to accurately identify and eliminate incorrect feature pair matches. Particularly under constrained spatial conditions, traditional map-based target localization methods such as Random Sample Consensus often fail to fully leverage geographical information and geometric relationships, hindering the accurate filtering of outliers. The above issue can lead to incorrect positional information, thereby reducing target localization accuracy. To address these challenges, we introduce the Local Affine Consistency Guided Triplet Sampling Consensus (LATSAC) algorithm, which is based on triplet feature matching. Benefit from a triplet-stable relationship, this method effectively filters outliers during UAV reconnaissance target localization missions in complex environments. To enhance the efficiency of the proposed algorithm in UAV reconnaissance missions, an improved initial data selection strategy is designed. In addition, to increase the target localization accuracy, a combined approach integrating a compensation term strategy with a data subset refinement strategy is developed. Experimental results from five challenging scenarios demonstrate that LATSAC achieves the lowest computational cost compared with that of existing state-of-the-art techniques for UAV reconnaissance target localization missions in complex environments. In addition, LATSAC exhibits superior outlier filtering performance when processing a high density of outliers, significantly improving target localization accuracy.

During the preliminary design of flight vehicles, i.e., missiles or guided rockets, propulsion system performance serves as a critical determinant of both maximum range and terminal velocity. However, complex grain configurations in solid rocket motors (SRMs) typically require geometric modeling software to obtain burning surface area, which severely constrains efficiency. To address this challenge, this study presents a neural network-enhanced rapid performance prediction and matching optimization framework for solid rocket motors (NN-SRM). In NN-SRM, neural networks are employed to simulate the evolution of key parameters during grain combustion, including burning surface area, grain volume, and moment of inertia. The zero-dimensional internal ballistics equations coupled with one-dimensional steady isentropic flow relations are incorporated into the framework to rapidly obtain thrust curves. A discrete–continuous mixed differential evolution algorithm is further employed to identify the optimal grain configuration that satisfies specific thrust requirements. Results demonstrate that, as for cylindrical, star, and finocyl grains, the neural network achieves R² exceeding 0.95. Finally, thrust matching optimization is conducted on three grains and achieves promising thrust solutions for the conditions of large thrust with short time and small thrust with long time, which demonstrates the effectiveness and practicality of the constructed NN-SRM.

Post-disaster Search and Rescue (SAR) missions demand rapid coordination of Heterogeneous Unmanned Aerial Vehicle (UAV) fleets under stringent payload and flight range limitations. Traditional heuristic solvers struggle to solve the Large-Scale Heterogeneous Team Orienteering Problem (LSH-TOP) within operational time limits due to the coupled complexity of task allocation and route planning. A Hierarchical Deep Reinforcement Learning framework decomposes this high-dimensional combinatorial problem into tractable sub-problems. An upper-level policy, guided by Monte Carlo Tree Search (MCTS), partitions the global target set to balance fleet workload distribution, whereas a lower-level Transformer-based model constructs near-optimal trajectories for individual agents. A Curriculum-Integrated Alternating Cooperative Training (C-ACT) protocol resolves the convergence difficulties associated with sparse feasible solutions in constrained environments. This protocol incorporates a dynamic constraint annealing strategy and a virtual agent buffer to progressively shape the solution space from relaxed to strictly constrained formulations. Experiments conducted on real-world geographic data demonstrate the proposed approach consistently outperforms all baselines across scales of 80 to 300 targets, improving over the strongest competitor by 0.63–8.51% and over conventional heuristics by up to 53.27% in objective value. Results indicate a task completion rate of 27.5% at the 300-target scale (versus 25.1% for the strongest baseline MCTS + OR) and balanced workload distribution, validating framework adaptability to complex emergency response scenarios.