基于滑动窗口CRITIC-G1的空战态势评估

传统的指标重要性相关(CRITIC)方法进行态势评估不能实现变权重,仅从数据本身属性出发进行赋权,忽略了专家意见,且CRITIC方法本身会使得图像抖动。针对以上问题,提出一种基于滑动窗口CRITIC-G1的近距离空战态势评估方法,将空战数据进行预处理,得到根据时间变化的优势函数矩阵,采用滑动窗口的方法将优势函数矩阵进行分解,通过对滑动窗口中的优势函数矩阵进行CRITIC-G1赋权,从而既实现变权重,又完成重要性排序的权重优化;同时引入变异系数降低综合优势函数图像抖动。最终输出飞机的综合优势函数值,对于两飞机的综合优势函数值进行分析得到态势评估的结果。仿真结果表明,滑动窗口CRITIC-G1法较传统CRITIC方法准确性提高了22.75%。

An Assessment of Situation in Air Combat Based on Sliding Window CRITIC-G1

Aimed at the problems that in the air combat, the traditional CRITIC (Criteria Importance Though Intercriteria Correlation) method is hard put to assess situation to achieve variable weights, assigning weights are only begging from the attributes of the data itself, expert opinions are neglected, and the CRITIC method itself makes image jitter, a sliding window CRITIC-G1 based close air combat situational assessment method is proposed. The method is to pre-processes the air combat data to obtain a dominance function matrix with change according to time, decompose the dominance function matrix by adopting the sliding window method, and assign CRITIC-G1 weights to the dominance function matrix in the sliding window, achieving both variable weights and the importance ranking-based weighting optimization. Simultaneously, the coefficient of variation is introduced to reduce the combined dominance function image jitter. The final output is the combined dominance function values of the aircraft, and the results of the situational assessment are obtained by analyzing the combined dominance function values of the two aircraft. The experimental results show that the sliding-window CRITIC-G1 method is an accuracy increase of 22.75%.