中国物理B ›› 2026, Vol. 35 ›› Issue (6): 68709-068709.doi: 10.1088/1674-1056/ae4f72
Environmental-gradient emotional memory memristive neural network circuit with TAP cell regulatory mechanism
Peng Qin(秦鹏)1, Tieqiao Liu(刘铁桥)2, Qiuzhen Wan(万求真)1,†, Rou Zhou(周柔)1, and Huaimin Xiang(向怀民)1
- 1 College of Information Science and Engineering, Hunan Normal University, Changsha 410081, China;
2 School of Information, Zhejiang University of Finance and Economics Dongfang College, Haining 314408, China
Environmental-gradient emotional memory memristive neural network circuit with TAP cell regulatory mechanism
Peng Qin(秦鹏)1, Tieqiao Liu(刘铁桥)2, Qiuzhen Wan(万求真)1,†, Rou Zhou(周柔)1, and Huaimin Xiang(向怀民)1
- 1 College of Information Science and Engineering, Hunan Normal University, Changsha 410081, China;
2 School of Information, Zhejiang University of Finance and Economics Dongfang College, Haining 314408, China
摘要： The hippocampus and amygdala in the human brain play a crucial role in the processing of emotion and memory. Specifically, the hippocampus encodes environmental information along with its corresponding emotional states, whereas the amygdala integrates emotional stimuli from the environment into the hippocampal memory system. Inspired by this biological mechanism and based on the principle of regulating cellular excitability within the mouse transient amplifying progenitor (TAP) mechanism, this study proposes an environmental-gradient emotional memory memristive neural network circuit. This bio-inspired neuromorphic circuit consists of two main modules: a hippocampal module and an amygdala module. The hippocampal module comprises two sub-modules: the environmental recognition neuron and the emotion generation neuron. The environmental recognition neuron is responsible for memorizing environmental features, while the emotion generation neuron establishes mapping relationships between the environment and emotional states. The amygdala module combines external emotional stimuli with internal current emotional stimuli to generate a comprehensive emotional assessment of the environment, and this emotional state can be stored. This memristive neural network circuit facilitates dynamic coupling between emotion and the environment, with the emotional output being continuous and graded rather than discrete. In PSPICE simulations, the proposed circuit exhibits satisfactory and stable functional performance. The findings of this study can offer valuable insights for the design of neuromorphic hardware circuits and for emotion simulation in bio-inspired robots.
中图分类号: (Models of single neurons and networks)
- 87.19.ll