战斗机突击作战实时辅助决策研究

Real-Time Assisted Decision-Making in Air Fighter's Penetration Attack

突击作战实时辅助决策对减轻飞行员负担、减少操作失误具有重要意义.在建立战机运动模型、威胁约束模型和目标攻击火控边界约束模型基础上,构建了突击作战辅助决策模型.引入了勒让德伪谱法,将辅助决策最优控制模型的解算转化为非线性规划求解问题.设计了基于滚动时域控制的优化策略,实现了模型解算的实时性.通过对基于C代码的可行序列二次规划(code feasible sequential quadratic programming,CFSQP)算法软件包的改进,有效减少了优化变量和约束条件数量,能对大规模非线性问题快速寻优.仿真结果表明了所提出的实时辅助决策方案的有效性和可行性.

Assistant decision-making is important in air fighter＇s penetration attack, especially for pilots to lighten flying burden and reduce operation mistakes. The air fighter kinematic and dynamical models, and the threat constraint and fire control border constraint models are built. Based on these models, an assistant decision-making model for air fighter＇s penetration attack is established. Using the Legendre pseudospectral method, the problem of calculating optimal control in assistant decision-making is converted to a nonlinear problem （NLP）. The receding horizontal control （RHC） method is used to ensure that the calculation can be complete in real time. By improving the code feasible sequential quadratic programming （CFSQP） algorithm, optimal results of the NLP problem can be quickly searched. Simulation studies show that the real-time assistant decision-making strategy works well.