Research On Threat Assessment Methods For Air Battlefield

With the rapid development of weapons and equipment,the air combat environment is increasingly complex and varied,and the modern battlefield presents highly intelligent characteristics.Threat assessment,as an auxiliary decision-making tool,can improve the real-time perception of commanders on the battlefield and help them judge the battlefield situation rapidly and accurately,so as to make scientific and effective tactical decisions.Based on the background of ground air defense,this thesis develops research on the issues related to air target threat assessment.The main research contents of the thesis are as follows:(1)For air attack targets,a threat assessment model is established.Many factors are considered in the modeling process,and threat assessment influencing factors are determined through correlation coefficients and Alpha coefficients.A threat assessment indicator system based on capability and situation factors was established,and the reasonableness of the model was verified with the actual cases.The situational information close to the real battlefield is obtained by means of simulation,and the threat value of the combination of dynamic and static is calculated by using the combination assignment method,and the threat value is quantified into threat levels.(2)Aiming at the problem that the penalty parameter (8 and the kernel parameter are difficult to set when using the Kernel Extreme Learning Machine(KELM)for air target threat assessment,the Multi-Strategy Based Improved Sparrow Search Algorithm(MISSA)is proposed.The method optimizes the Sparrow Search Algorithm(SSA)in four aspects: Tent chaotic backward learning initialization strategy,nonlinear inertia weights,nonlinear global search strategy and adaptive t-distribution,and conducts comparison experiments in nine typical benchmark functions.The experimental results show that the MISSA algorithm has superior performance in terms of convergence speed and accuracy compared with the other six heuristics.(3)For large scale battlefield data,a MISSA-KELM based threat assessment algorithm was established.In order to evaluate the algorithm more accurately and reasonably,a 10-fold crossvalidation was used in the training process.Analyzing the experimental results,the Precision of MISSA-KELM is 0.946,Recall is 0.947,and F1-score is 0.947,which are all better than IPSOKELM,IGA-KELM,ISOA-KELM,and HGWO-KELM,and the results show that the MISSAKELM algorithm has significant advantages.In summary,the thesis builds a threat assessment model based on information warfare and focuses on the complex military requirements in the intelligent battlefield.For the problem of large scale of battlefield data and more deceptive data,threat assessment based on KELM and its improved algorithm is proposed to improve the performance of threat assessment algorithm.