Ordered Time-Sensitive Target Allocation Strategies Based on Deep Learning

Aiming at the problem of “order-based” target allocation for time-sensitive targets, we analyze the situation in depth, establish the target optimization model under different requirements. Firstly, based on the posture information, a graph theory model is established to synthesize and analyze. Then, based on the known conditions, we establish the target damage model, cost consumption model, and mission action process time model under the conditions of autonomous order bidding, and then we get the order condition judgment model of strategy <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$A$</tex> and B. Considering the cost consumption and mission action process time constraints comprehensively, we solve and visualize the two kinds of order bidding scenarios in the main text. Finally, an objective optimization model is established according to the requirements of minimizing cost and maximizing the probability of destruction under strategy <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$A$</tex>, and a depth-high priority hierarchical adaptive branching (DHAB) algorithm with lower spatial and temporal complexity is proposed for solving the most optimal requisitioning scheme. In addition, the DHAB is iterated using the chaotic quantum particle swarm algorithm with adaptive crossover operator (ACCQPSO) with cost and kill probability as the optimization objectives, respectively, to solve the dispatching scheme under the conditions of minimizing the cost and maximizing the damage under strategy <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{A}$</tex> in an end-toend manner, and obtain the minimum cost and the maximum average kill probability.