Transient Chaotic Neural Networks Based on Sinusoidal Feedback Terms and Their Applications
Bin Zhou, Nan Xu · 2023
There is a novel chaotic neural network model that uses sinusoidal perturbations to add nonlinear elements to enhance the chaotic properties, and then employs maximum Lyapunov exponents and neuronal inverse bifurcation maps to modify the model parameters to ensure that the network has chaotic dynamic properties. Simulation data demonstrate the feasibility of the scheme and how changing the value of the has an impact on the coefficients in the excitation function and the self-feedback term on the output of the model, either by increasing the range of the chaotic traversal or by enhancing the speed of convergence of the chaotic traversal, reflecting the sensitivity of the chaotic neural network model to the initial values of the parameters. The experimental findings show that the transient chaotic neural network model based on sinusoidal perturbations is capable of handling the complexity and diversity of the combinatorial optimization problem and can successfully find the global optimal solution when the new model is applied to a combinatorial optimization problem (TSP problem). providing a new solution to the combinatorial optimization problem.