Abstract: The two-wheeled self-balancing vehicle is a class of non-complete system of intrinsic instability. Its complex nonlinear dynamical equations include many characteristics, such as higher order, strong coupling, less driving. This paper uses Newtonian mechanics to model the two-wheeled self-balancing vehicle, and constructs an adaptive mechanism by feedforward, feedback and inputs. Based on them, this paper presents a model reference adaptive control（Model Reference Adaptive Control, MRAC） strategy to control the posture and speed of the two-wheeled self-balancing vehicle. Simulation results show that the MRAC algorithm can ensure the stability of the system and make the system approach to the dynamic performance of ideal reference model. Meanwhile it makes the system obtain favorable robustness near the equilibrium point.

Key words: two-wheeled self-balancing vehicle, model reference adaptive control, robustness

摘要： 两轮自平衡小车是一类非完整的本征不稳定系统，其动力学方程具有复杂非线性、高阶次、强耦合、欠驱动等特点，采用牛顿力学分析法进行系统建模，利用前馈、反馈以及输入量构成自适应机构，在此基础上提出模型参考自适应控制（Model Reference Adaptive Control，MRAC）策略对两轮自平衡小车的姿态和速度进行控制。通过仿真，结果表明采用MRAC算法能够在保证系统稳定的前提下，获得接近于给定理想参考模型的动态性能，并使系统在平衡点附近具有良好的鲁棒性。

关键词: 两轮自平衡小车, 模型参考自适应控制, 鲁棒性