基于卡尔曼滤波的旋转倒立摆智能控制算法研究

发布时间：2018-01-14 14:40

本文关键词：基于卡尔曼滤波的旋转倒立摆智能控制算法研究　出处：《重庆理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：基于经典、现代及智能控制理论的各种控制策略实现了倒立摆系统的稳摆控制、轨迹跟踪控制,这些控制算法均需以摆杆的角度和角速度信号作为稳定控制的依据,但研究者设计控制算法时多未考虑各种算法实现过中难免要受系统噪声、量测噪声、外界干扰和建模误差的影响,比如实时控制中摆杆的角度信号会受到测量噪音污染,在获取角速度时该噪音被放大,导致控制器的输出有较大的波动,又如电机执行过程中也会混入过程噪音使系统运行不稳定,故这样设计出的控制器鲁棒性差、抗外界扰动能力有限,而将卡尔曼滤波算法与控制算法进行组合设计应用于提升倒立摆系统稳定控制器、轨迹跟踪控制器性能的研究少有人涉及,故本文选择旋转倒立摆为控制对象,对卡尔曼滤波算法在提升控制器控制性能方面进行了下面的研究工作:(1)考虑旋转倒立摆系统实际运行中,会存在机械磨损、各种摩擦、外界干扰、内部干扰等建立了旋转倒立摆非线性动力学标称模型和线性状态空间模型。(2)应用LQR控制器对旋转倒立摆线性状态空间模型进行稳定控制,并通过仿真实验观察其控制效果,仿真结果显示LQR控制器在含有各种噪声的旋转倒立摆系统应用中,难以获得精确的输出反馈量实现对系统的最优控制,故引入卡尔曼滤波算法设计了状态观测器对LQR控制器进行改善,仿真结果显示方法有效。(3)考虑到旋转倒立摆的非线性特性,设计了基于无迹卡尔曼滤波的LQR控制器,进一步提高LQR控制器的控制精度和鲁棒性,并设计了仿真实验验证方法的有效性和可行性。(4)针对旋转倒立摆的非线性动力学标称模型,提出了变增益滑模控制策略进行稳定控制和轨迹跟踪控制,并通过仿真实验观察控制效果。(5)针对变增益滑模控制器仿真出现的问题,提出了基于无迹卡尔曼滤波的非奇异终端滑模控制策略,并设计了仿真实验,仿真结果验证了方法的有效性和优越性。
[Abstract]:Based on the classical, modern and intelligent control theory, various control strategies are used to realize the stability control and trajectory tracking control of inverted pendulum system. These control algorithms need to take the angle and angular velocity signal of pendulum rod as the basis of stable control, but the researchers design the control algorithm without considering the system noise and measurement noise in the realization of all kinds of algorithms. The external interference and modeling error, such as the angle signal of the pendulum rod in real-time control, will be polluted by measurement noise, and the noise will be amplified when the angular velocity is obtained, resulting in a large fluctuation of the output of the controller. For example, the noise will also be mixed in the process of the motor to make the system run unstable, so the controller designed in this way has poor robustness and limited ability to resist external disturbances. The combination of Kalman filter algorithm and control algorithm is applied to improve the stability controller of inverted pendulum system. The research of trajectory tracking controller is seldom involved, so this paper chooses the rotating inverted pendulum as the control object. The following research work is done on the Kalman filter algorithm to improve the control performance of the controller. (1) considering the actual operation of the rotating inverted pendulum system, there will be mechanical wear, various kinds of friction and external interference. The nonlinear dynamic nominal model and linear state space model of rotating inverted pendulum are established. The linear state space model of rotating inverted pendulum is controlled by LQR controller. The simulation results show that the LQR controller is difficult to obtain accurate output feedback in the application of rotating inverted pendulum system with various noises to achieve the optimal control of the system. Therefore, the Kalman filter algorithm is introduced to design the state observer to improve the LQR controller. The simulation results show that the method is effective and takes into account the nonlinear characteristics of the rotating inverted pendulum. The LQR controller based on unscented Kalman filter is designed to further improve the control accuracy and robustness of the LQR controller. Finally, the effectiveness and feasibility of the simulation experiment are designed. (4) aiming at the nonlinear dynamic nominal model of the inverted pendulum, a variable gain sliding mode control strategy is proposed for stability control and trajectory tracking control. And through the simulation experiment observation control effect. 5) aiming at the problem of variable gain sliding mode controller simulation, a non-singular terminal sliding mode control strategy based on unscented Kalman filter is proposed, and the simulation experiment is designed. Simulation results show the effectiveness and superiority of the method.
【学位授予单位】：重庆理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN713;TP273

【参考文献】