Human-robot Interactive Task Parsing Method Based on Large Language Model

Traditional mobile robots need specific task delivery format or follow certain rules to perform tasks, which greatly reduces the user’s experience and the freedom of use. In order to solve the problem of large restriction in traditional task execution, we designed an interactive task parsing method based on large language model, aiming to provide an interactive more flexible and reliable task parsing method suitable for mobile robots. We have carried out experiments to verify the accuracy and flexibility of the method.