【作者】 周艳；

【导师】 南海鹏；

【作者基本信息】 西安理工大学 ， 水利水电工程， 2004， 硕士

【摘要】 自抗扰控制理论通过对系统未知扰动和不确定量的估计，并将估计量反馈回控制系统中，从而实现系统的动态线性化，该理论思想独特，方法简单，特别适合于具有非线性和时变特性的水轮发电机组这样的系统。本文研究自抗扰控制理论在水轮发电机组中的应用，为设计对强非线性和不确定强扰动都能保证很高的控制精度的可靠的水轮发电机组控制策略作理论上的支持。 在分析非线性跟踪—微分器(TD)输入输出特性的基础上，提出了跟踪时间不随输入信号幅值改变的TD。针对水轮发电机组的非最小相位特性，基于设置过渡过程的思想，在传统的PID控制器的基础上应用改进的TD设置合理的过渡过程，有效的减小了非最小相位中的反调。使之用于功率设定值变化的水轮发电机组有功功率控制中，仿真结果表明了该方法的有效性。 分析了自抗扰控制器和其中扩张状态观测器(ESO)的特性，用神经网络补偿控制对象的部分变化，从而减轻扩张状态观测器的负担，获得了“鲁棒性”更强的基于神经网络的自抗扰控制器。针对水轮发电机组的非线性和不确定性，把神经网络与自抗扰控制理论相结合，并依据对象特性对控制器做了结构调整。通过仿真比较，证实了基于神经网络的自抗扰控制器能对参数变化范围更广的水轮发电机组进行有效控制。更多还原

【Abstract】 Through the unknown perturbation and uncertain estimation of the system, and feedback the estimation to the control system, the active disturbance rejection controller thus realizes the dynamic feedback linearization of those dynamic systems. The ADRC theory is unique, and the method is simple, it suited to the non-linear an uncertain hydroelectric power sets.Have analyzing the tracking-differentiator(TD) I/O characteristic, to get a improving TD which output uninfluenced by the input. For the non-minimum phase characteristic of hydraulic turbine power set, this paper proposes and analyses the tracking-differentiator (TD) with invariable tracking time based on the idea of arranging the transient process. The rational transient process is set by the improving TD based on the classical PID controller, which effectively reduces undershoot of non-minimum phase system. The simulation result shows that it is efficacious when applied to the hydraulic turbine power set real power control.Have analyzing the characteristic of ADRC and ESO and compensate some changes of target by artificial neural networks to lighten the burden of ESO, thus get an stronger controller which base on ANN and ADRC. For the nonlinear and uncertain characteristic of hydraulic turbine power set, this paper combines the artificial neural networks (ANN) with ADRC theory and adjusts the structure of the controller according to the characteristic of the subject. The simulation result shows that the ADRC, which base on ANN are efficacious when applied to a hydraulic turbine power set which change largely and quickly.更多还原