Exploring Physics-Informed Neural Networks for the Generalized Nonlinear Sine-Gordon Equation

Algorithm 1

PINN algorithm for NLSGE.

	Require: Training data, collocation points , contains interior and boundary points.
	Initial condition, boundary condition, and the NLSGE.
(1)	Define network architecture (input layer, hidden layers, output layer, activation function, and optimizer).
(2)	Initialize weights and biases , .
(3)	for all epochs do
(4)	apply forward propagation:
(5)	compute the residual:
(6)	compute loss:
(7)	apply the optimizer: .
(8)	end for