A Design And Implementation For Two-wheeled Self-balancing Mobile Platform

The development of MEMS and integrated circuit technology promote wheeled mobilerobots’ perception system and decision-making system replacement. The wheeled mobilerobots have been in the booming development while also have faced the more complexapplication environment and the mare tasks to be processed.Compared with multi-wheel mobile robot, two wheeled mobile robot has significantadvantages. In the first place, the structure of these robots is more concise. Secondly, theserobots are lighter and small. Thirdly, two wheeled robot has much smaller radius of gyration,even more can be achieved turning in place with any turning radius. Fourthly, the two mobilerobots only need strong forward drive capability with a certain angle can easily leap obstacles.Given these advantages, we design a non-manned, two wheeled self-balancing mobileplatform. It can be self-balancing, and do motion control according to the user instructionsfrom the host computer. The platform has good scalability and robustness. The mobileplatform with intelligent sensors and processors can be set to the navigate robot which can beused in large-scale science and technology museum, supermarkets, hospitals, banks, museumsand other places.The mobile platform we designed is a basis platform of two self-balancing service robot.It can automatically detect the current tilt angle. According to the inclination angle and anglevariation rate, it regulates the output torque so as to achieve the purpose of dynamic balancing.The data and interactive instruction between mobile platform and the host computer arecommunicated through the communication protocol self-designed. Through communicationprotocols, users can monitor platform’s real-time status in running, and can also operateplatform’s motion via software in PC.Mobile platform’s torque control is achieved by current PI closed loop control whichcontrols brushless DC motors. In current PI closed-loop control, we design the FIR filter forfiltering the feedback current. This enables more precise output torque, while ensuring therobustness of PI closed-loop control.The paper uses Kalman filter algorithm to achieve two self-balancing mobile platform’stilt angle’s accuracy. And we design experiments to verify the accuracy and effectiveness of the Kalman filter algorithm.Paper combines fuzzy control theory and PID control technology, then proposes a fuzzyPID control theory to control two self-balancing mobile platform’s motion.Two wheeled self-balancing mobile platform in this paper has two operating modes, inwhich the remote control mode is accomplished by means of the PC. The communicationprotocol designed between PC with mobile platforms is rather special, and its data packet’schecksum state machine is designed more comprehensive, which can effectively reduce thedata packet’s error rate.