Optimalization of Neural Network Method on Harmonis Damped Swing

Abstract

Numerical methods are used to solve the differential equations in the case of a damped pendulum system. In this paper, we aim to solve the equation of dumped pendulum motion using some numerical. Four numerical methods are decomposed, each the Euler method, fourth-order Runge-Kutte, Odeint provided in Python, and Neural Network scheme. All answers obtained are approximations or predictions that include errors. The error of those methods will be compared with the analytical solution of the case, known as the global error. The Odeint and fourth-order Runge-Kutta methods are more accurate than the other methods. The Odeint is built from the Runge-Kutta fourth-order method.The Neural Network method has less accuracy than Runge-Kutta, but the error obtained is still within acceptable tolerance limits.