几类非完整与欠驱动系统的控制研究
发布时间：2018-08-05 15:57
【摘要】：在实际系统中,非线性、时滞和外部随机干扰等因素是广泛存在的,传统的分析方法因为不能准确地对各种非线性不确定性因素进行建模,容易造成控制性能的下降甚至失败.因此,针对非线性系统领域的研究显得尤为重要.本文针对几类非完整和欠驱动系统中存在的非线性问题进行了研究,设计了相应的控制器并进行了稳定性分析,主要研究成果包括:1.对一类带有未知参数的欠驱动柔性关节机器人系统进行了研究,由于现存的研究成果均不同程度地具有过度参数化的特点,增加了控制器设计的复杂度.本文找到引起上述问题的原因,研究它的自适应跟踪控制问题,主要解决了以下问题:(i)减少过度参数化带来的复杂度,利用一个参数实现跟踪控制.(ii)利用Backstepping控制方法找到一个合适的Lyapunov函数,设计了控制器,并证明了所设计的控制器能够使得闭环系统全局稳定.2.对一类带有不确定参数和时滞的非完整系统,研究了它的状态反馈镇定问题.首次考虑了带有多个时滞状态的不确定非完整系统的自适应控制问题,引入了时滞非线性漂移的边界条件,利用基于动态增益的Lyapunov-Krasovskii方法和技巧,设计并构造了相应的状态反馈自适应控制器,并证明了所设计的控制器能够使得闭环系统全局渐近稳定.3.对一类带有未知系数和外界干扰的链式非线性系统,研究了它的固定时间跟踪问题.利用增加幂次积分法将有限时间跟踪问题变换为固定时间跟踪问题,所设计的控制律能够保证系统状态在固定时间跟踪到期望轨迹.此类带有快速响应和短时切换的固定时间镇定问题在实际应用中具有非常高的理论意义和应用价值.4.对一类典型的欠驱动非完整桥式吊车系统进行了研究,此类系统具有高度的耦合性和非线性,分析了基于理论模型的仿真结果同实际运行结果存在差异的主要原因,通过引入RBF神经网络自适应控制算法对非线性不确定因素进行逼近,并结合确定学习方法设计了基于RBF的神经网络自适应控制器,实现了针对实际吊车系统RE3DCrane的控制和应用,所设计的控制器鲁棒性较好且有一定的实用价值.
[Abstract]:In the practical system, the factors such as nonlinearity, time-delay and external random disturbance exist widely. Because the traditional analysis method can not accurately model all kinds of nonlinear uncertain factors, it is easy to cause the degradation or even failure of control performance. Therefore, the study of nonlinear systems is particularly important. In this paper, the nonlinear problems in some nonholonomic and underactuated systems are studied, the corresponding controllers are designed and the stability is analyzed. The main research results include: 1. A class of underactuated flexible joint robot systems with unknown parameters is studied. Due to the overparameterization of existing research results, the complexity of controller design is increased. In this paper, the causes of the above problems are found, and its adaptive tracking control problem is studied. The main solutions are as follows: (i) reduces the complexity caused by excessive parameterization. The tracking control. (ii) is realized by one parameter. The Backstepping control method is used to find a suitable Lyapunov function. The controller is designed, and it is proved that the designed controller can make the closed-loop system globally stable. 2. The state feedback stabilization problem for a class of nonholonomic systems with uncertain parameters and delays is studied. The adaptive control problem of uncertain nonholonomic systems with multiple time-delay states is considered for the first time. The boundary condition of time-delay nonlinear drift is introduced, and the Lyapunov-Krasovskii method and technique based on dynamic gain are used. The corresponding state feedback adaptive controller is designed and constructed, and it is proved that the designed controller can make the closed-loop system asymptotically stable globally. The fixed time tracking problem for a class of chain nonlinear systems with unknown coefficients and external disturbances is studied. The finite time tracking problem is transformed into a fixed time tracking problem by adding power integration method. The designed control law can ensure that the state of the system can be tracked to the desired trajectory at a fixed time. This kind of fixed time stabilization problem with fast response and short time switching is of great theoretical significance and practical value in practical applications. A kind of typical underactuated non-holonomic bridge crane system is studied in this paper. This kind of system is highly coupled and nonlinear. The main reasons for the difference between the simulation results based on the theoretical model and the actual operation results are analyzed. By introducing RBF neural network adaptive control algorithm to approximate nonlinear uncertain factors and combining with deterministic learning method, a neural network adaptive controller based on RBF is designed. The control and application of RE3DCrane for actual crane system are realized. The controller designed has good robustness and practical value.
【学位授予单位】：曲阜师范大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：O231
本文编号：2166272
[Abstract]:In the practical system, the factors such as nonlinearity, time-delay and external random disturbance exist widely. Because the traditional analysis method can not accurately model all kinds of nonlinear uncertain factors, it is easy to cause the degradation or even failure of control performance. Therefore, the study of nonlinear systems is particularly important. In this paper, the nonlinear problems in some nonholonomic and underactuated systems are studied, the corresponding controllers are designed and the stability is analyzed. The main research results include: 1. A class of underactuated flexible joint robot systems with unknown parameters is studied. Due to the overparameterization of existing research results, the complexity of controller design is increased. In this paper, the causes of the above problems are found, and its adaptive tracking control problem is studied. The main solutions are as follows: (i) reduces the complexity caused by excessive parameterization. The tracking control. (ii) is realized by one parameter. The Backstepping control method is used to find a suitable Lyapunov function. The controller is designed, and it is proved that the designed controller can make the closed-loop system globally stable. 2. The state feedback stabilization problem for a class of nonholonomic systems with uncertain parameters and delays is studied. The adaptive control problem of uncertain nonholonomic systems with multiple time-delay states is considered for the first time. The boundary condition of time-delay nonlinear drift is introduced, and the Lyapunov-Krasovskii method and technique based on dynamic gain are used. The corresponding state feedback adaptive controller is designed and constructed, and it is proved that the designed controller can make the closed-loop system asymptotically stable globally. The fixed time tracking problem for a class of chain nonlinear systems with unknown coefficients and external disturbances is studied. The finite time tracking problem is transformed into a fixed time tracking problem by adding power integration method. The designed control law can ensure that the state of the system can be tracked to the desired trajectory at a fixed time. This kind of fixed time stabilization problem with fast response and short time switching is of great theoretical significance and practical value in practical applications. A kind of typical underactuated non-holonomic bridge crane system is studied in this paper. This kind of system is highly coupled and nonlinear. The main reasons for the difference between the simulation results based on the theoretical model and the actual operation results are analyzed. By introducing RBF neural network adaptive control algorithm to approximate nonlinear uncertain factors and combining with deterministic learning method, a neural network adaptive controller based on RBF is designed. The control and application of RE3DCrane for actual crane system are realized. The controller designed has good robustness and practical value.
【学位授予单位】：曲阜师范大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：O231
【参考文献】
相关期刊论文 前4条
1 孙宁;方勇纯;;一类欠驱动系统的控制方法综述[J];智能系统学报;2011年03期
2 高丙团;陈宏钧;张晓华;;龙门吊车系统的动力学建模[J];计算机仿真;2006年02期
3 王朝立,霍伟;用滑动模态实现一类非完整动力学系统的指数镇定[J];自动化学报;2000年02期
4 华克强,高淑玲,朱齐丹;吊车防摆技术的研究[J];控制理论与应用;1992年06期
相关博士学位论文 前1条
1 孙宗耀;高阶不确定非线性系统的控制设计和性能分析[D];山东大学;2009年
，本文编号：2166272
本文链接：https://www.wllwen.com/shoufeilunwen/jckxbs/2166272.html
