A Memristor-Based Neural Network Circuit With Latent Inhibition and Transient Forgetting Effects and Application in Industrial Intelligent Grasping
Junwei Sun, Yijin Shen, Peng Liu, Yanfeng Wang · IEEE Transactions on Industrial Informatics · 2024
Conditioning and associative memory play an important role in the learning process of biological brain, and many neural network circuits have been designed to reproduce the relevant classical experiments. However, these circuits are mainly devoted to the realization of various phenomena in the acquisition process, such as reacquisition, generalization, differentiation, and blocking. A Pavlov associative memory circuit based on memristor is designed in this article. The circuit realizes latent inhibition effect and a variety of biological forgetting features. The correctness of implementing these functions described above is demonstrated by simulation results in Pspice. This circuit can realize the influence of external environment changes on the learning and forgetting states of organisms. It offers valuable insights for modeling biological intelligence and simulating the learning function of human associative memory. Particularly, associative memory and transient forgetting can be applied to industrial intelligent grasping robots.