| With the development of science and technology, mobile robots are becoming more and more popular in people’s life. The application field of mobile robots has gradually shifted from military, industry, shopping malls and hospitals to home domestic usage. In the future, domestic intelligent mobile robot will be an essential part of people’s life. Higher and higher level of artificial intelligence is required to be implemented in future mobile robots. The ability of indoor positioning and auto-charge is an important factor to evaluate the level of artificial intelligence. Only when the robots can know their exact position in the room can they move automatically. After obtaining the exact position in the room, the mobile robot can move towards the charging home and accomplish the connection with the charging home when it detects that the battery level is low.In order to reduce the hardware cost and calculation complexity as well as improve positioning accuracy, photoelectric image sensors are employed in our work as the positioning sensors. This thesis will first introduce the basic principle of photoelectric image sensors. The hardware circuit of the photoelectric image sensor will be presented along with the equation of linear regression obtained by software programming and experiments. Secondly, we will divide the mobile robot movement into three categories:transition, rotation and avoiding obstacles. A mathematical model will be established and a detailed positioning algorithm will be elaborated for each movement. The existing problems and their corresponding solutions will also be discussed. Utilizing photoelectric image sensors for indoor robot positioning not only reduce the hardware cost but also improve the accuracy of positioning, because the error of each measurement is very low and the errors of two consecutive measurements can be cancelled out.After the robot moves to the area near the charging home by its positioning information, the connection program will start in order to accomplish the connection between the robot and the charging home. In the implementation of the connection system, we design the method of infrared connection. The robot can judge its relative position to the charging home by detecting the different information sent by three infrared transmitters on the charging home. Then the robot can adjust its position accordingly until it connects to the charging home. The infrared connection system has the advantage of low hardware and software implementation cost as well as high connection rate and low time consumption.According to research work and experimental results above, this thesis makes a deep investigation of the key technology of the photoelectric image sensor and its application in robot indoor positioning. The experimental results show that the positioning error is less than1%on indoor wooden floor, which satisfy the demand of indoor robot positioning. The connection rate of the infrared connection system is above98%and its time consumption is85seconds at most, which meet the demand of auto-charge as well. The research results of this thesis have important value in the field of mobile robot indoor positioning and auto-charge. |
PDF Full Text Request