Memristor-Based Neural Network Circuit of Emotional Grading in Intelligence Decision System

Emotion is a crucial factor in neuromorphic computing. Current memristive circuits have explored the impact of emotion on operant conditioning, there is a lack of behavioral research on emotional decision-making. This paper proposes a memristor-based neural network circuit of emotional grading in intelligence decision system, which mainly includes emotion generation module, emotion grading module, emotion monitoring module and intelligent decision-making module. This circuit can convert external stimulus signals into emotional signals and classify the emotional signals for processing. The graded emotional signals are used to make decisions through a memristive neural network, and ultimately output the behavior after the decision. The emotional intensity will also be monitored in real-time to ensure that it is within the normal range. The simulation results are validated by PSPICE, providing a reference for artificial intelligence and brain like computing.