Aiming at the dynamic game and confrontation characteristics between the offensive and defensive sides in the strike process of time-sensitive targets, this paper constructs a game model for the decision-making of both sides based on evolutionary game theory, and solves their evolutionary stable strategies. Numerical simulation is adopted to investigate the effects of factors such as the initial probability of strategy selection and target threat degree on the strategic choices of both sides. The results show that the initial probability of strategy selection cannot change the final stable state of bilateral strategy selection, but only affects the strategy convergence speed. Target threat degree, the quantity ratio of true and false targets, and the cost difference of strategies exert significant influences on the strategy selection of the offensive side with an obvious threshold effect. These factors have a certain impact on the strategy convergence rate of the defensive side, while they do not alter its final stable state.
This paper proposes a flight early warning modeling and path assurance methodology for weather modification UAV missions. A spatiotemporally continuous three-dimensional meteorological field is constructed based on regional weather forecasting models, incorporating critical atmospheric variables—including wind speed, humidity, and supercooled water—into the operational environment. Environment modeling is further refined by integrating topographical data, risk-zone constraints, and icing avoidance mechanisms. Concurrently, mission-specific and safety constraints are established to facilitate robust path planning under complex meteorological conditions. Experimental results demonstrate that the proposed method is capable of generating flight trajectories that optimize the balance between operational requirements, target reachability, and flight safety within a coupled 3D terrain-meteorology environment.
To address the lack of hierarchical analysis methods for combat networks, this paper proposes HD-OCLPA, an operational chain based label propagation algorithm for multi-hierarchical detection. The method identifies hub nodes via degree centrality, uses label propagation to form initial communities, then extracts operational chains through breadth-first search and iteratively aggregates communities based on a shared chain matrix to construct a clear hierarchy tree. Experiments show that community layers vary significantly across network scales, indicating that internal community organization, rather than scale alone, determines hierarchical complexity. Hub nodes play a key role in aggregation, and community merging reveals the evolutionary path of combat units from decentralized execution to integrated application. This study provides theoretical support for vertical command hierarchical structure design.
To address problems such as the unclear connotation of software general quality characteristics and the lack of dedicated detection and evaluation methods, this paper designs and implements a software-testing-based evaluation framework for software general quality characteristics. Grounded in the basic theory of general quality characteristics, the framework clarifies the fundamental connotation of software general quality characteristics and establishes a corresponding basic indicator system. It also develops a basic model set for evaluating software general quality characteristics on the basis of general utility-evaluation techniques such as Sigmoid and Gauss functions. By embedding and integrating the evaluation of software general quality characteristics into the software testing process, a testing-based intensive work mode for evaluating software general quality characteristics is proposed. Engineering practice shows that, without changing existing work procedures, the proposed work mode can support the detection and evaluation of software general quality characteristics primarily through software testing. It therefore provides a feasible and effective solution for evaluating software general quality characteristics solely on the basis of software testing, without conducting dedicated special-purpose tests.
Addressing the hub node location problem for multi-echelon oil reserve networks, this paper constructs an integer programming model aimed at cost minimization, considering multiple military constraints such as transportation logic, demand satisfaction, receiving and dispatching limits, flow balance, load balancing, and timeliness. To tackle the discreteness, multiple constraints, and fragmented feasible regions of the model, an Improved Aptenodytes Forsteri Optimization (IAFO) algorithm is proposed. The algorithm achieves accurate mapping from continuous to discrete spaces through dual-priority encoding and feasible-domain-guided decoding, integrates adaptive inertia weight search with Lévy flight and population disturbance-based jumping search to balance exploitation and exploration, and incorporates elite opposition-based learning and restart mechanisms to suppress premature convergence. Case study results demonstrate that compared with genetic algorithm, particle swarm optimization, and the original AFO, IAFO reduces the total cost of the obtained solutions by an average of 4.58%, 2.15%, and 1.37%, respectively, while achieving superior solution stability and decision consistency. The methodological framework established in this paper provides effective theoretical and technical support for military oil reserve location decisions.
