【作者】 王鑫；

【导师】 潘天红；

【作者基本信息】 安徽大学 ， 控制科学与工程， 2022， 硕士

【摘要】 自抗扰控制(Active Disturbance Rejection Control,ADRC)由于不依赖被控对象的精确模型,只需知道系统的相对阶次,便能够通过系统的输入输出信息来对被控系统的内部状态和总扰动进行估计和补偿,所以具有很强的适应性,并在无人机、航天器、电机等诸多领域得到广泛应用。然而ADRC的控制效果与被控系统的初始状态、ADRC参数密切相关。为此,本文在深入分析ADRC结构的基础上,对线性扩张状态观测器(Linear Extended State Observer,LESO)的结构进行优化,给出一种改进型的ADRC(improved ADRC,i ADRC),并借助与内模控制(Internal Model Control,IMC)在二自由度结构下的等效关系,提出i ADRC的参数整定方法。最后以高阶柔性直线系统为被控对象,对提出的算法进行验证。主要研究工作如下:在深入分析ADRC结构和原理的基础上,分析了ADRC各模块的主要功能以及相关参数对性能的影响。然后对LESO的结构进行剖析,研究发现,在非零初始状态下,系统输出的初始观测值和实际值偏差过大会导致初始峰值现象。为此,将在传统LESO中引入了新的校正变量,以提高观测器的观测精度,同时有效地抑制初始峰值现象的影响。然后,对IMC和i ADRC两种控制器进行结构分析和转化,发现两者在二自由度结构下可以近似等效,通过等效之后可以证明i ADRC的两个带宽值与IMC中的两个时间常数互为倒数。因此可以通过IMC参数来对i ADRC进行参数整定,随后通过仿真验证了其有效性。最后,以高阶柔性直线系统作为实际被控对象,建立了系统的数学模型,并将系统内部建模不确定性、外界未知干扰和其他不确定因素集总起来定义为广义总扰动;通过设计状态观测器和控制器,对被控系统的状态变量和总扰动进行估计和补偿。与PID、传统LADRC相比,改进的自抗扰控制器具有精度高、抗扰性好和抑制初始峰值现象等优点。更多还原

【Abstract】 Active disturbance rejection control can estimate and compensate the internal state and total disturbance of the controlled system through the input and output information of the system without depending on the precise model of the controlled object,only knowing the relative order of the controlled system,so it is widely used in many fields such as UAV control,spacecraft attitude control,permanent magnet synchronous motor control and so on.However,the control effect of ADRC is closely related to the initial state of the controlled system and parameters.Therefore,on the basis of in-depth analysis of ADRC structure,this paper optimizes the structure of linear expansion state observer,gives an improved ADRC,and puts forward a set of parameter tuning methods by virtue of the equivalent relationship with internal model control in the two-degree-of-freedom structure.Finally,the high-order flexible rectilinear system is taken as the controlled object to verify the algorithm.The research work is summed up,as follows:On the basis of in-depth analysis of the structure and principle of ADRC,the main functions of each module of ADRC and the influence of related parameters on performance are analyzed.Then,the structure of LESO is analyzed,and it is found that in the non-zero initial state,the deviation between the initial observation value and the actual value is too large,which leads to the initial peak phenomenon.Therefore,a new correction variable will be introduced into the traditional LESO to improve the observation accuracy of the observer and effectively suppress the influence of the initial peak phenomenon.Then,the structural analysis and transformation of IMC and i ADRC controllers are carried out,and it is found that they can be approximately equivalent under the two-degree-of-freedom structure.After equivalence,it can be proved that the two bandwidth values of i ADRC and the two time constants in IMC are reciprocal.Therefore,i ADRC can be tuned by IMC parameters,and then its validity is verified by simulation.Finally,taking the high-order flexible rectilinear system as the actual controlled object,the mathematical model is established,and the internal modeling uncertainty,external unknown disturbance and other uncertain factors are aggregated to define generalized total disturbance.By designing the state observer and controller,the state variables and total disturbance of the controlled system are estimated and compensated.Compared with PID and traditional LARC,the improved ADRC has the advantages of high precision,good disturbance rejection and suppression of initial peak phenomenon.更多还原