【作者】 娄宇；

【导师】 吴爱国；

【作者基本信息】 天津大学 ， 控制理论与控制工程， 2009， 硕士

【摘要】 锻造液压机由于锻造速度快、控制精度好、自动化程度高、工艺范围广以及节能、节材显著,被认为是锻造设备的发展方向之一,是目前大型自由锻锤及中小型普通水压机的更新换代产品。随着现代工业的发展,要求液压机可以实现多对象控制、多轴同步控制,超低速运行,高精度定位,以及实现整体的联合控制。本文归纳总结了国内外锻压机及其控制系统的发展现状及趋势,讨论了液压机的工作原理。阐述反步递推算法基本概念。在深入分析等温锻造液压机系统的运行原理的基础上,建立了液压机四缸系统整体数学模型。该模型是一个高阶仿射非线性模型,变量的类型分别为压力、位置和速度,所有变量在实际系统中均可观测,同时液压机四缸之间速度、压力存在耦合关系,而且由于负载力的不断变化,该系统模型参数具有不确定性。在建模的过程中,本文充分考虑了负载力变化对液压机四缸同步的影响,建立了更加真实的仿射非线性模型。针对该模型,在李亚普诺夫稳定性理论基础上,创新的将反步递推算法推广到高阶多输入多输出的仿射非线性系统,并且在仿真中得到了很好的控制效果。应用该算法,在液压机压力能力范围内,当负载力发生变化时,依然可以实现对任意速度的准确跟踪,并实现液压机四缸的位置准确同步。针对系统内负载力的剧烈变化,首次设计高阶自适应反步控制器,将参数自适应性算法融入反步递推设计之中,推导出参数自适应率,该算法对参数的变化有较强的适应性。仿真结果表明,自适应反步控制器相对于直接反步控制器,在速度和位置跟踪的同时,具有更强的抗干扰性能。自适应反步递推设计不需要对模型线性化或者解耦处理,根据模型的结构直接对模型设计控制器,且对参数扰动适应性很强,在模型变量的可控范围内,可对任意给定输入实现准确跟踪。在模型结构准确的基础上,该算法对参数变化适应性较强,且算法在实际中易于实现,对于锻造液压机的同步控制具有良好的控制效果。更多还原

【Abstract】 Forging hydraulic press is considered as one of the developing directions of forging press because of its fast forging speed, high control accuracy, high automation, wide technology range, saving energy and material. It is a product to replace large type of forging hammer and mesotype and minitype common forging water press. Because of the improvement of industry, we need liquid press can control multi-axis and through simple configuration, we can use intelligent method to make the result more precise.In the thesis, the current status and development tendency of hydraulic forging press domestic and abroad are investigated. The work principle of hydraulic press and its control system are studied. Based on the analysis of isothermal hydraulic forging press, the mathematical model of the hydraulic system is established. The model with multi-input and multi-output is high-rank. The variable is position, speed and pressure. And all variables all can be measured in the control system. During working out the mathematical model, I calculate the influence between the changing of load and synchronization of four vats, so the mathematical model is closer to reality. Base on Lyapunov thesis, the backstepping is used for the model and the result of simulation is very good. In the control system, the load will be changing, but under the ability of the machine, the output can track the reference precisely and realize the synchronization of four vats. When the load changes sharp, the adaptive backstepping is used for the changing model. The algorithm is more adaptive for changing parameters. The result of simulation indicates that adaptive backstepping controller is more robust and steady.During the design of adaptive backstepping controller, we don’t need linearization and anti-coupling for the model, so we design the controller directly. And the control system is more robust and under the ability of the machine, the output can track the reference precisely. Base on the precise configuration of the model, the algorithm is realized easily, and the result is better than the backstepping system.更多还原