Abstract: The rapid and accurate evaluation of electromagnetic environment complexity is the basis of situation and threat assessment. The static evaluation method has been studied in the previous work. On this basis, the various suitable indexes for dynamic evaluation are discussed and optimized, and a novel neural networks is proposed though combining the cloud model and constructive neural networks. Then a dynamic evaluation method of electromagnetic environment complexity is proposed. The objectivity and accuracy are improved in the proposed evaluation method. The factual radio monitoring data experiment results indicate the availability and practicability of this method.

Key words: electromagnetic environment complexity, constructive neural networks, cloud model

摘要： 战场电磁环境的快速准确估计是态势和威胁评估的前提。在电磁环境复杂度静态评估的基础上，优化了适合动态评估的指标集，将云模型与构造型神经网络相融合为构造型云神经网络，提出基于构造型云神经网络的电磁环境复杂度动态评估方法。该方法能够提高评估结果的客观性以及准确率，无线电监测数据验证了方法有效性。

关键词: 电磁环境复杂度, 构造型神经网络, 云模型