Hybrid Automaton

Hybrid automata are mathematical models used to represent systems exhibiting both discrete and continuous behavior, aiming to bridge the gap between symbolic and numerical reasoning. Current research focuses on improving the efficiency and robustness of algorithms for learning and verifying hybrid automata, including advancements in active automata learning techniques that handle noisy or conflicting data and methods for synthesizing models from time-series data. These advancements are significant for applications such as robotics, where accurate modeling of physical systems with discrete control actions and continuous dynamics is crucial for tasks like planning and verification of safe robot behavior.