Design And Control Research Of The Robot Riding Bicycle System

At the time of various kinds of mobile robots industry are flourishing and mature,the bicycle robot whose two wheels are lengthways arrangement can be rapidly through the long and narrow space.It also has the advantages of simple structure and low-energy consumption,but it gets few significant development in the society and academia.This article direct at the bicycle robot is a typical symmetry asymmetric under-actuated nonholonomic constraint system which is difficult to realize its balance control.It is expected to control the motor speed and acceleration according to the quality of bicycle body and its geometric relations,through the derivation and analysis of dynamic model,by the torque equation and force balance,which can drive the inertia-wheel rotation to maintain the bicycle balance under different roll angle.At the same time,it is driven by multi-joint humanoid robot‘riding' the bicycle,so that the bicycle robot can realize effective self-balancing rectilinear,simple curve and combination path riding.The main works of this paper lie in the bicycle robot dynamics modeling and the balance controller design,which includes exploring two kinds of bicycle robot system's relationship between bicycle roll angle with the inertia wheel motor output value and the bicycle roll angle with the steering angle during the dynamic steering.In this paper,the design will dynamics model by the Newton Euler method,employ the traditional dual-loop PID controller for the inertial-wheel motor feedback control from the system of ‘robot riding bicycle',take the multiple-servos cooperative control instructions to control the multi-joint humanoid robot riding motions and the handlebar servo and realize wireless communication,the obstacle avoidance and riding on the predetermined trajectory by Bluetooth module,ultrasonic,and reserve control instructions by time series.At the same time,direct at the simplified bicycle robot system,the further design will take Lagrange-Eulerian method to dynamics model and employ the fuzzy adaptive PID controllers.Through the simulation,the two schemes' feasibility and control effect are verified and tested.In the end,theentity prototype experiments and the analysis of ‘robot riding bicycle' system are conducted and get a favorable control effect.At present,the domestic research is mainly about the bicycle robot without mechanical auxiliary structure.This kind of mobile robot need to steer the handlebar constantly in the process of running towards.It can't get through a narrow space by a low speed and realize precise trajectory control.It also needs to match the scheduled path in real time and has the poor robustness.Its mathematical model is complex and needs huge computation.Under the funding of Exhibits Key Technology Research and Development Project,Robot Technology Interactive Experience Series from National Science-technology Support Plan,the system of ‘robot riding bicycle' this paper studied has the advantages of small and simple mechanism,low-cost,strong anti-interference.And it lays a certain foundation for the other forms of bicycle robot,unicycle robot and other under-actuated system balance control in future.