A neuronal circuit based on a second-order memristor

With the increasing advances in computer science, a neuron with information memory is needed to form an artificial neural network. In this paper, a novel non-volatile second-order memristor (SOM) is applied to an RLC neuron to simulate synaptic connections and electromagnetic radiation, thus constructing a Second-Order Memristor RLC (SOM-RLC) neuron and giving a analog circuit. Firstly, numerical methods are used to analyze the changes in neuronal dynamic behavior caused by internal and external magnetic fields and synapses, including various bifurcation behaviors and extreme multi stability. Considering that changes in magnetic field can cause energy flow in neurons, the energy distribution of neurons is analyzed. Finally, by constructing an analog circuit of SOM-RLC neuron and implementing chaotic attractors on DSP, the data results are verified, proving its engineering feasibility. Analyzed the biomimetic performance of SOM-RLC neuron from different perspectives, explained the main characteristics of neurons under internal and external magnetic field interference, and further advancing the exploitation of artificial neurons.