论文部分内容阅读
首先通过时间加权末位淘汰机制对最小二乘支持向量机(LS-SVM)算法进行改进。其次,将这一算法应用于导引头视线被云层遮挡(穿云)或其他原因导致导引头失锁时对视线角速率的预测。导引头处于锁定状态时应用该算法进行在线训练,导引头视处于失锁状态使用训练形成的决策函数对视线角速率进行在线预测。最后,通过弹道末端设置导引头失锁的数学仿真结果,统计使用预估视线角速率(决策函数输出)作为末端导引信息的多条弹道的脱靶量,确认了最小二乘支持向量机对典型视线角速率信号预测的有效性和用于提高小型空地战术导弹穿云和抗干扰能力的应用前景。
Firstly, LS-SVM algorithm is improved by time-weighted elimination mechanism. Secondly, this algorithm is applied to predict the angular rate of line-of-sight when the seeker’s vision is blocked by clouds (clouds) or other reasons. When the seeker is in the locked state, the algorithm is applied to online training, and the seeker is in the out-of-lock state to predict the line-of-sight angular rate online using the decision function formed by the training. Finally, the mathematical simulation results of the lost-seeker of the seeker at the end of the ballistic trajectory are used to calculate the miss distance of multiple trajectories using the estimated line-of-sight angular velocity (output of the decision function) as the end guidance information. It is confirmed that the least square support vector machine The validity of typical line-of-sight angular rate signal prediction and the application prospects for improving the ability of small space tactical missiles to penetrate clouds and resist interference.