Mechanics-informed neural networks (MINNs) are a new neural networks concept. These networks base their structure on physical (mainly mechanics) and control theory laws.
The framework's goal is to allow the users fast modeling and control of a mechanical system such as an autonomous vehicle, an industrial robot, a walking robot, a flying drone.
A conceptual representation of the mechanical system will be used to obtain a neural network of the MINN type, which has an optimal structure to model the considered mechanical device. The framework will realize the training by appropriately choosing all the hyper-parameters and allowing adequate training by providing suitable experimental data. The framework will allow the user to synthesize and train a neural network that will be used as a control system in a few simple steps and without the need to perform new experiments. The realized controller will be exported independently from any external library in the C language.