Design Of Two-wheel Self-balancing Robot Based On BLE Indoor Positioning And Navigation

Nowadays,people's demands for indoor positioning is more and more intense because of the development of mobile communication technology and the change of people's living habits.However,due to the complex environment,low positioning accuracy,small coverage area and other factors,engineers(who major in mobile communication)still take indoor positioning technology as one of the key research objectives.In this paper,Bluetooth Low Energy(BLE),which has the advantages of low power consumption,small size,wide coverage and high precision,is used as the main tool(of research)for indoor positioning.And we take twowheeled self-balancing robot with small size,small diameter,low cost,and flexibility as an experimental platform.With combining indoor positioning technology and self-balancing robot technology,the research for two-wheel self-balancing robot based on BLE which is used in indoor positioning and navigation is realized.In this paper,firstly,Kalman filter algorithm is used to analyze the optimal attitude signal of gyroscope and accelerometer.Then PID control algorithm is used to realize the balance control of the two-wheel self-balancing robot,and PD control serves as a position closed-loop system to realize the vertical function of the two-wheeled self-balancing robot.And PI control serves as a speed closed-loop system to realize the forward and backward movement.The hardware circuit system of two-wheel self-balancing robot and BLE beacon node(which provides a stable working platform for indoor positioning)was designed by the circuit diagram drawing software---Altium Designer.According to the experimental environment and the results of multiple tests,we put BLE beacon nodes and select experimental sampling nodes reasonably and efficiently,which improve the positioning accuracy without increasing the cost on experiment.The offline fingerprint database is established according to RSSI and NAME.RSSI data packets of surrounding beacon nodes are collected by the same data acquisition method in the real-time positioning stage.The real-time position of the two-wheeled selfbalancing robot was determined by comparing the fingerprint database and using the indoor localization algorithm---k-nearest neighbor algorithm.Finally,the indoor positioning and navigation of the self-balancing robot is realized through path planning by reading the Angle data of the gyroscope.In this paper,amount number of experiments and simulations by MATLAB show that the self-balancing robot can maintain the system balance stably and achieve their own forward and backward and left and right steering function.In terms of indoor positioning algorithm,compared with traditional WKNN algorithm,this paper proposes a new WKNN positioning algorithm with an adaptive K value.According to the actual positioning environment and the RSSI data packet of beacon node collected at the anchor point,the K value is changed adaptively to obtain the positioning result with the highest accuracy,which provides the basis for further design of experiment.