针对多功能雷达(Multi-Function Radar,MFR)状态预测方法存在的鲁棒性、预测正确率不佳的问题,提出一种基于机动特征辅助的MFR状态预测方法.该方法将载机机动信息与常规侦收参数共同作为预测特征集,一方面利用支持向量回归(Support Vector Regression,SVR)和侦收信号特征集,得到常规预测模型,另一方面通过SVR和机动特征集,得到MFR各个状态间的转变概率模型;然后利用D-S证据理论得到最终预测状态.实验结果表明,与SVR和LSR方法相比,平均预测精度分别提高了6.97%和7.2% ,同时具备更优异的鲁棒性.此外,提出的预测方法通过进一步的拓展,可应用于机械设备、道路交通等领域.%Aiming at the problem of weak robustness and poor accuracy of the traditional multi-function radar ( MFR) state prediction methods,a MFR state prediction method aided by maneuvering features is proposed. First of all,the aircraft maneuver features and the conventional reconnaissance parameters work together as the prediction feature set;then, on the one hand,the conventional prediction model can be achieved by support vector regression (SVR) and detected signal feature set;on the other hand,the MFR state transition probability model of each state is obtained by SVR and maneuvering feature set;at last,the eventual prediction state is obtained by the D-S evidence theory. Experimental results show that in comparison with SVR and LSR,the proposed method improves the average prediction accuracy by 6.97% and 7.2% re-spectively,and meanwhile,it is more robust. The proposed method can be applied to the mechanical equipment,road trans-port,etc. by further extension.
展开▼
