Research Of Intelligent Multi-function Wheelchair

As the aging population is expanding fast, the quality of life of older persons have begunto enter the public view. The intelligent wheelchair, as an auxiliary tool for the older anddisabled, has received a large number of research and gain great developments. Intelligentwheelchair, as one wheeled mobile robot, combines the technology in many fields, forexample, navigation and obstacle avoidance technology and human-computer interactiontechnology. The intelligent wheelchair must have the safer and more reliable system and moreconvenient, diverse operational method.This paper utilizes a series of methods to achieve motion control system, navigationsystem and human-computer interface system of intelligent wheelchair. This paper utilizespath planning algorithm based on3-splines to avoid obstacle. The paths planned by3-splines method have the property of3-order geometric continuity and can keep the smoothand steady in motion.In order to achieve the function of visual track and navigation of wheelchair, this paperproposed the algorithm integrates visual track method which based on Kinect camera andobstacle avoidance and navigation method. And the proposed algorithm has achievedsatisfying results. This paper designs two control methods to improve the operational diversityof intelligent wheelchair. The two control methods are Brain-Computer-Interface controlmethod based motion imagery and control method based speech recognize, respectively. Inthe design of Brain-Computer-Interface based motion imagery, Common Space Patternsalgorithm is used to extract the feature in EEG signal. And support Vector Machine methodclassifies these features to recognize the human’s EEG signal and to control the wheelchairaccording human’s intention. This paper proposed a control method combines the voice detecttechnology with Google Speech recognize engine. The voice detect technology is to detect thebeginning and the end of human’s voice, then post the valid voice data to Google Speechrecognize engine. The results can be control the wheelchair by comparing with the controlcommands.An electric wheelchair is modified into an intelligent wheelchair according the functionabove. And a lot of experiments have been conducted, such as the obstacle avoidanceexperiments, visual track and navigation experiments, brain control experiments and voicerecognize experiments. The results illustrated in this paper show that the algorithms andstrategies above has achieved satisfying performance. And the intelligent wheelchair wedesigned is feasible and practical.