Research And Design Of Multi-agent Software Model For Modular Robot Dog

The multi-agent system is a focal point of research in the field of distributed artificialintelligence. MAS originated in imitation of the human forms of social organization, and theMAS consists of different agents which can work independently to achieve local tasks andcollaborate to achieve global tasks. MAS can make the system with greater flexibility andadaptability, and better able to adapt to the open, dynamic environment. So MAS subjects to theattention of more and more people, and also provides a new approach to the study and design ofintelligent robot software system.MAS is of great significance and actual using value in the field of intelligent robots. Inorder to obtain a better application with MAS, we build an embedded robot dog platform andpropose an intelligent robot model, relying on the National Natural Science Foundation of China(61175094). The platform has walking mechanism, many kinds of perceptual (visual, speech,infrared, temperature etc.) as well as wireless communication modules, and the features ofinteraction, learning and so on. And we design and divide robot into few agents by the idea,which is “thinking and learning, perception and interaction, action and control”，so the softwarearchitecture is flexible and not depend on the specific hardware. What’s more, the model takesaccount of both intelligence and response capability for emergency by deliberative module basedon CBR and reflex module.After simulating on the JADE platform, we design the Speech Recognition and CBRsoftware in the interactive and learning platform, and design the UART communication protocolsand control software in the robot dog.With the software, robot dog is able to walk, avoidobstacles, perceive and process external signals such as voice input and infrared. The modelprovides reference value for the design of embedded intelligent robots. In future research, wewill improve and optimize the model and apply it to the design of other embedded intelligentrobots, highlight the CBR reasoning mechanism and the instantaneity of the reasoning process.