In the acoustic analysis of mechanical systems that produce high sound pressures, the nonlinearity of sound waves must be considered. Under the conditions of nonlinearity, a change in the speed of sound distorts the waveform. This waveform distortion is important in the design of acoustic equipment. Numerical analysis is commonly used to analyze nonlinear sound waves. Recently, there have been many reports on the use of machine learning for numerical analysis. However, in data-driven approaches, the governing equations may not be satisfied by simulation results. To address this problem, we present a new approach for the analysis of nonlinear sound waves using Physics-Informed Neural Networks (PINNs). In this study, a PINN algorithm for nonlinear acoustic equations was presented and a one-dimensional traveling wave was analyzed. The results obtained were consistent with the nonlinearity of sound waves. Considering that PINN is a mesh-free numerical analysis method, and that trained PINN performed acoustic analysis faster than FDM, PINN is expected to be a new option for nonlinear acoustic wave analysis.