Model And Control Of An Unmanned Bicycle

From the perspective of control, bicycle is an unstable system, subjects to non-holonomic contact constraints when rolling without slipping. Stabilizing an unmanned bicycle proves to be an intractable problem, it has been attracting considerable attention in the field of automation. This article uses the method of modeling, analyzing, integrating, designing and realizing, studies the control of balance for an unmanned bicycle.First, this article presents the inverted pendulum model and multi-body model of a bicycle at different assumptions. It also analyzes bicycle dynamic behaviors, including non-minimum phase steering behavior of the front fork, and stability of the system according to velocity. Taken velocity as the fuzzy variable, a kind of fuzzy T-S model is presented based on multi-body model, and local linear theory is adopted to design a fuzzy controller, which offers a new method to maintain balance of the system under velocity changes.Second, based on bicycle dynamic behaviors, two generations of unmanned bicycle control system are developed, which realize the function of regulating velocity, turning the front fork, sensing attitude information of the bicycle; the advantage and disadvantage of them are compared, too. Master-slave control is adopted. Based on attitude information of the bicycle which is acquired by MCU, a computer calculates the optimal steering angle and sends it to MCU, MCU drives the actuator after receiving it.Last, some experiments are done to study the control of the unmanned bicycle. In order to suppress the noise, Butterworth low-pass digital filter is used to preprocess the sampling signal in real time. It proves that turning the front fork has great influence on the stability of the system. A PID control algorithm is designed, and it can keep balance of the system in some time, which is the groundwork for completely control of an unmanned bicycle.