The rapid advancement of deep neural networks has opened new avenues for understanding complex physical systems. This master’s thesis explores the application of deep neural networks in determining the dynamics of physical systems. The research objectives include the development of new methodologies for modelling and predicting the behaviour of dynamic systems, and harnessing the power of artificial intelligence. To achieve these objectives, we investigated the principles of deep learning and neural networks. Deep neural networks were trained and adapted using real data from physical systems, enabling a relatively accurate characterization of system dynamics. The results demonstrated the effectiveness of this approach in modelling simple physical systems and its potential for application on more complex systems. Key findings include the ability to capture both simple and more complex patterns in system behaviour, approaching the level of precision and adaptability characteristic of conventional modelling techniques. The implications of this research extend to various domains, including physics, engineering, and data-driven decision-making. In conclu sion, this master’s thesis showcases the potential of deep neural networks as tools for determining the dynamics of physical systems. Applying artificial intelligence to such problems promises to enhance our understanding of complex systems and facilitate data-driven solutions.