Research Of Obstacle Avoidance Of Intelligent Wheelchair Based On Ultrasonic

With the development of society, the senior citizen is in course of aggravating; further more because of the road accident frequently happen, the handicapped persons are also increasing with each passing day. How to provide advantageous walking implement to these who have difficulty on walking, to improce their mobility and make them to be in harmony with society has become a important problem. Applying moving robot technology to electric wheelchair ,intelligent wheelchair have just been raised recently.Since the intelligent wheelchair is the serving robot which help the people who has difficulty in walking, its design proposal is to let people feel comfortable and to avoide obstacle safely is also a integrant function. This project is exactly to bring forward to implement this function, the content of this paper is as follows:1.The design of ultrasonic obstacle avoidance systemOn the basic of the analysis of intelligent wheelchair function, the paper modularize the design proposal of intelligent wheelchair and divide it into machinery structure part, man-machine interface part, controlling module part , the driving part and environment perception part .Among them, environment perception part and obstacle avoidance system is discussed particular. Behind comparing a lot of sensors, we used ultrasonic sensor as the environment perception component, the circuit chart which is used to receive and emit the ultrasonic signal is designed here .How to dispose sensors is also been discussed.2.Wheelchairs build model in kinematicses and distance analysis of obstacle avoidance safelyAfter analysis of wheelchair moving characteristic, we get the connection function between right wheel and left wheel of the wheelchair when it is moving and calculate the analysis of distance and speed which wheelchair can avoid obstacle safely.3. Data inosculate of multi- ultrasonic sensorAfter the detailed analysis of the fuzzy control and the neural network, the algorithm of the data inosculate of the multi-ultrasonic sensor based on the fuzzy neural network is designed. The BP network is optimized by the method of the descending momentum gradient. The paper brought forward a simple efficient arithmetic for ultrasonic to make the calculation simple. After fuzzy the inosculate distance data,we classify the obstacle environment and differentiate it using optimizing BP network, finally we get the information which is used to avoid obstacle safely.