禁忌学习神经元模型的电路设计及其动力学研究

Circuit design of tabu learning neuron models and their dynamic behavior

对禁忌学习混沌神经元模型进行了详细的电路设计,包括单神经元、具有线性相似函数的双神经元以及具有二次相似函数的双神经元模型,并运用EWB(Electronic Workbench)电路仿真软件对所设计的电路进行了仿真,研究了电路中的Hopf分叉和混沌等非线性动力学现象,通过与数值仿真结果的比对验证了所设计电路的正确性与合理性.

The circuit design and implementation of neuron models have attracted much attention in recent years due to its importance in both theoretical studies and applications. In this paper,we designed electronic circuits for the recently proposed tabu learning chaotic neuron model,including the circuit design of single tabu learning neuron,a two-neuron system with linear proximity function,as well as a two-neuron system with quadratic proximity function. We used the electronic workbench （ EWB） software to perform simulations of the designed circuits. We also studied the nonlinear behavior,especially the Hopf bifurcation and chaos,of the designed circuits. The consistency between the nonlinear behavior in the designed circuits and that in the numerical simulations demonstrates the correctness of the circuits design.