变负载伺服系统自适应内模控制研究

【作者】 吴杰；

【导师】 郭毓；

【作者基本信息】 南京理工大学 ， 控制理论与控制工程， 2009， 硕士

【摘要】 负载是伺服系统的重要组成部分,负载大范围的变化将会影响伺服系统的控制性能,甚至导致系统不稳定。本文主要讨论了自适应控制和内模控制方法在变负载伺服系统中的应用,进行了相关理论分析和仿真研究。本文首先分析了变化负载是如何影响伺服系统的,通过仿真,进一步分析了负载在不同变化情况下对伺服系统动态性能、稳态性能和系统稳定性的影响。然后根据变负载伺服系统的特性设计了模型参考自适应(MRAC)控制器,仿真结果表明,MRAC在负载慢变情况下的控制效果较好,而在负载跃变时控制效果欠理想。为提高MRAC对快时变负载的适应能力,提出了并行自适应PID控制策略,控制器由MRAC控制器和常规PID控制器综合而成,仿真结果验证了并行自适应PID控制效果比MRAC有明显的改善。针对伺服系统负载变化的问题,本文还设计了伺服系统的内模(IMC)控制器,仿真结果表明,IMC在模型匹配时可以取得较好的控制效果,而在模型失配时控制性能较差。针对模型失配情况下IMC控制性能降低的情况,进一步设计了变负载伺服系统的自适应内模控制器,控制器以内模控制为基础,通过自适应机构在线辨识出对象模型和其逆模型,仿真结果表明,该算法在负载慢变和跃变情况下都可以获得较好的控制性能。更多还原

【Abstract】 Load is one of the most important components of a servo system, whose large-scale change will affect the control performance of the servo system and even lead to its instability. The application of adaptive control and Internal Model Control in servo system with variable load is discussed.Meanwhile the theoretical analysis and the simulation results are presented in this thesis.The problem about how the changing load affects servo systems is studied firstly in this paper, and further, the effects on dynamic performance, steady-state performance and system stability of servo system by changing load is analyzed.A Model Reference Adaptive Controller (MRAC) is designed according to the characteristic of Variable Load Servo System. Simulation results shows good performance achieved when the load varies slowly while lower performance achieved when load varies fast using MRAC. A parallel adaptive PID control strategy consisting of MRAC and PID controller is proposed to enhance the fast time-varying adaptive capacity of MRAC.The simulation results show that the parallel self-adaptive PID controller has better performance than MRAC.An Internal Model Controller for the issue of changing load in servo system is designed.Simulation results show that the IMC has better performance in the case of model matching than model mismatching. Since the control performance of IMC is low when the model mismatches, the adaptive Internal Model Controller is designed.The controller based on IMC uses the adaptive institutions to identify the object model and its inverse model online.The simulation results show that this algorithm has better performance in both slow-varying and hopping load conditions.更多还原