论文部分内容阅读
在文献[1]提出的模糊形态联想记忆网络FMAM的基础上,提出了一种新型的模糊形态学双向联想记忆网络FMBAM,证明了FMBAM的双向联想中能够保证记忆在一步之内完成,因此不存在收敛问题,也表明了FM-BAM具有优越的抗腐蚀或膨胀噪声的能力。但是,通常的噪声是随机的,为此,本文提出了动态核的方法,从而较好地提高了FMBAM对随机噪声的抗噪能力。仿真实验验证了利用动态核的双向联想记忆网络FMBAM。在加入较大的随机噪声的情况下,仍能保证完全记忆在一步之内完成。
Based on the fuzzy shape associative memory network FMAM proposed in [1], a novel fuzzy morphological bi-directional associative memory network FMBAM is proposed, which proves that FMBAM bidirectional associativity can ensure that the memory is completed in one step and therefore not There are convergence problems, but also shows the FM-BAM has excellent corrosion resistance or expansion of the noise capacity. However, the usual noise is random. Therefore, this paper presents a dynamic kernel method, which can improve the noise immunity of FMBAM to random noise. Simulation experiments verify the use of dynamic nuclear bidirectional associative memory network FMBAM. In the case of adding a large random noise, can still guarantee complete memory in one step to complete.