Network-based Control System Of Meal-assistance Robot

A network-based control system of Meal-assistance Robot is designed. Internet and Controller Area Network is used to achieve the target of collecting information from all kinds of sensors as well as sending the calculation result. Two computer based on x86 architecture, one for video possessing and the other for control algorithms writing in Matlab/Simulink blocks, is used as central information processing computer connected with Internet. The center research work focus on the hardware & software design for Meal-Assistance Robot and its experimental research.Firstly, the paper introduced the background and significance of Meal-Assistance Robot as well as the main research content according to development status of home and abroad. The main challenge and approaching strategy of system design is analyzed through mechanical structure and function demand. The implement method of hardware & software is explained. Video processing computer and xPC target computer is of the two main part of the control system. Video processing computer is responsible for video acquisition, identifying algorithms, sending the calculation result and displaying the graphical user interface. xPC target computer which consist of the interface module and real-time control computer is responsible for sensor signal acquisition, realization of control algorithms and logics, the output of processing result.This paper has given a detailed discussion on thought, corresponding algorithms and implement method. A hardware structure of communication node based on industry ARM and CPLD is designed. Several kinds of functional module including analogue to digit, digit to analogue, digital signal input and output; encoder collection is constructed on the communication node. The embedded software based onμC/OS-Ⅲreal time operation system kernel realized the compilation and download of the language draw by Simulink module cooperated with the Matlab interface software using xPC target. Human face expression mode achieved by robot control program and template-based facial expression recognition algorithm realized the food taken, fetching and feeding function by robotic arm. A total test including hardware & software is included. The results show that the performances of the network-based Meal-assistance robot meet the requirements and achieve the design task.