| A wheeled mobile intelligent robot experiment platform is designed in this thesis, based on which a deep and systematic study has been done about person recognition and tracking and obstacle avoidance.First of all, all the design schemes of the wheeled mobile robot platform are given, includes mechanical structure, hardware circuit and control software, etc. In design of the platform, power system, transmission system, materials selection, chassis design and structure in mechanical structure scheme, total structure, wireless communication circuit, motor control circuit, data acquisition circuit and GPRS control circuit in hardware circuit scheme and software structure, obstacle avoidance process and information acquisition process in control software scheme are given separately.Secondly, an approach used to recognize person is proposed. Several methods are presented of the mobile robot platform, includes video acquisition and pre-treatment, person recognition based on feature, a fast and practical method to remove object shadow, etc. And the given recognition results illustrate that the proposed approach for person recognition is satisfied with the request of wheeled mobile robot platform in order to track person.At last, the approach using vision information and ultrasonic information, simulation and experiments for mobile robot platform to track person and the method combining person tracking and obstacle avoidance strategy are given separately. Through introducing closed-loop feedback theory of control theory into the process of object tracking, a novel algorithm is proposed, which is named a multi-layer sub-PID control algorithm. Simulation and experiments results illustrate that the function of tracking person is realized in the wheel-like robot platform. |
PDF Full Text Request