多特征融合与深度学习的防空目标分配方法

Air defense target assignment method based on multi-feature fusion anddeep learning

为了提升多平台协同防空作战条件下大批量目标分配的快速性和准确性,分析了当前目标分配算法的特点与局限,提出了基于多特征融合与深度学习的防空目标分配思路和建模方法,通过选取多个影响目标分配效果的关键特征属性作为输入,构建了适应特征规模的深度神经网络目标分配优化模型结构。经仿真验证,该方法相比典型方法在进行目标分配时的拦截效能更好,耗时更短,能够更全面地描述战场环境特征因素,平衡优化深度神经网络目标分配模型的过拟合和欠拟合问题,有效提高目标分配的计算速度和准确性。

In order to improve the rapidity and accuracy of mass targets assignment under the condition of multi-platform cooperative air defense engagement,the features and limitations of current target assignment algorithms are analyzed,and a design idea and modeling method of air defense target assignment method is proposed based on multi-feature fusion and deep learning.By selecting multiple key feature attributes that affect the target assignment as input,the structure of target assignment optimization model is constructed based on deep neural network adapted to the scale of feature.Comparing with typical methods,the simulation results show that this proposed method can achieve better effectiveness on target assignment issues,and can describe the characteristic factors of battlefield environment more comprehensively,the over fitting and under fitting problems of the target assignment model based on deep neural network are balanced and optimized,and the calculation speed and accuracy of target assignment are effectively improved.