事件触发机制下线性NCS主—被动混合容错控制研究

发布时间：2018-08-06 22:06

【摘要】：网络化控制系统(Network Control System,NCS)是一种通过有限的数字通信网络在传感器、执行器和控制器之间进行信息传输的空间分布式系统。在许多实际的NCS中,由于种种原因系统模型总是存在着不确定性,而信息在网络中传输又不可避免地出现时延、丢包等问题,作为控制系统的组成部件,执行器或传感器也随时可能发生各种故障,导致控制系统的性能下降甚至不稳定,因此将容错控制技术引入到NCS的安全性和可靠性的研究中日益受到人们的广泛关注。容错控制主要分为被动容错控制(PFTC)和主动容错控制(AFTC),PFTC是利用鲁棒控制技术使系统对集内故障不敏感,但是由于系统正常和故障时控制器是同一个控制增益,保守性较大。AFTC方法是针对在线估计故障结果重组或重构新的控制器,由于在线估计故障和控制器的重组或重构需要时间,因此AFTC实时性较差。离散事件触发机制是给定事件触发条件,通过判断该条件成立与否决定信息是否传输,而网络诱导时延产生的根本原因是有限的网络带宽,引入离散事件触发机制可以明显降低通讯负载维持系统稳定。目前针对事件触发机制下的NCS容错控制的研究主要集中在PFTC以及滤波器的设计。由此为了有效的节约网络资源并结合主动和被动的优缺点设计基于事件触发机制的混合控制器就可能成为人们研究的主要方向。鉴于此,本文针对具有时变时延的线性不确定NCS,基于事件触发机制,研究了执行器任意失效故障情形下的主被动混合鲁棒容错控制问题。总体概略如下:1)基于事件触发机制研究标称线性NCS的主被动混合容错控制问题针对具有时变时延的标称线性NCS,首先分别基于独立事件发生器和带有事件发生器的智能传感器(集成事件发生器),建立故障系统模型,设计使系统在发生故障集以内故障时稳定的被动容错控制器。同时设计故障诊断观测器估计任意执行器失效故障的大小,一旦获得准确的故障信息立即重构控制器以补偿故障的影响,最后通过仿真对比独立事件发生器和集成事件发生器下数据传送量,为后续在离散事件触发机制下的主被动混合鲁棒容错控制研究奠定了基础。2)基于事件触发机制研究线性不确定NCS的主被动混合鲁棒容错控制问题针对具有时变时延的线性不确定NCS,建立基于离散事件触发通讯机制的NCS故障模型,并基于H_∞控制思想,设计了鲁棒故障检测观测器;离线设计被动鲁棒容错控制器确保系统在已知故障发生时稳定,未知故障发生初期减缓系统性能下降速度;同时利用鲁棒故障检测观测器实时在线检测故障,并重构控制器补偿任意未知故障对系统的影响。3)基于事件触发机制研究线性不确定NCS主被动切换鲁棒H_∞容错控制问题针对具有时变时延的线性不确定NCS,在受到外部有限能量扰动的影响下,基于离散事件触发通讯机制,分别设计正常控制器和被动鲁棒H_∞容错控制器,使系统正常运行时能具备良好的动态性能,当系统发生故障时,通过瞬态切换函数平滑切换至被动鲁棒H_∞容错控制器,确保系统在已知故障发生时不仅稳定而且具有一定的H_∞扰动抑制性能,未知故障发生初期减缓系统性能下降速度;同时设计鲁棒H_∞故障检测观测器,实时检测故障,利用自适应补偿控制消除未知故障对系统的影响。4)在以上研究的基础上,采用仿真实例对所有系统的被动容错控制器、主动容错控制器、混合容错控制器和切换函数设计的结果分别进行了仿真研究以及系统性能的影响分析,其结果表明文中所得结论是正确有效的。对于上述研究结论,仿真结果表明,在传感器端引入事件发生器,在节约网络资源的情况下,实现了具有时变时延线性不确定NCS的故障估计与主被动混合鲁棒容错控制集成设计的研究。所提出的基于观测器的故障诊断方法可以有效的诊断任意故障,在系统运行中所发生的不同情形,采用状态反馈控制策略,设计的主被动混合鲁棒容错控制器,使得系统无论是在正常运行状态还是发生执行器任意故障不但渐近稳定都具有良好动态性能。
[Abstract]:Network Control System (NCS) is a spatial distributed system that transmissions information between sensors, actuators and controllers through a limited digital communication network. In many practical NCS, there are always uncertainties in the system model for various reasons, and the transmission of information is unavoidable in the network. As a component of the control system, as a component of the control system, the actuator or sensor may also have various faults at any time, causing the performance of the control system to decline or even unstable. Therefore, the fault tolerant control technology is widely concerned in the research of the security and reliability of NCS. Fault tolerant control is the main problem. It is divided into passive fault-tolerant control (PFTC) and active fault-tolerant control (AFTC). PFTC is the use of robust control technology to make the system insensitive to intra set faults, but because the controller is the same control gain when the system is normal and fault, the conservativeness.AFTC method is to restructure or reconstruct a new controller for the online estimation of the fault results. The reconfiguration or reconfiguration of the fault and controller takes time, so the real-time performance of AFTC is poor. The trigger mechanism of the discrete event is the trigger condition of a given event. By judging whether the condition is established or not, the transmission of information is determined, and the root cause of the generation of network induced delay is the limited network bandwidth, and the trigger mechanism of the discrete event can be obvious. To reduce the communication load to maintain the stability of the system, the research on NCS fault tolerant control under the event triggering mechanism is mainly focused on the design of PFTC and filter. In order to effectively save network resources and combine the advantages and disadvantages of active and passive, the design of hybrid controller based on event triggering mechanism may become the main research owner. In view of this, this paper aims at linear uncertain NCS with time-varying delay. Based on event triggering mechanism, this paper studies the passive hybrid robust fault-tolerant control problem in the case of arbitrary failure of the actuator. The overall outline is as follows: 1) the problem of the active and passive hybrid fault-tolerant control based on the event triggering mechanism to study the nominal linear NCS The nominal linear NCS of time-varying delay is based on the independent event generator and the intelligent sensor with event generator (integrated event generator). The fault system model is established and the passive fault-tolerant controller is designed to stabilize the system when the fault is within the fault set. The fault diagnosis observer is used to estimate the loss of the arbitrary actuator. The size of the effective fault, once the accurate fault information is obtained, the controller is reconstructed immediately to compensate for the effect of the fault. Finally, the data transmission is compared between the independent event generator and the integrated event generator by simulation, and the base.2 based on the event is laid for the subsequent study of the passive hybrid robust fault-tolerant control under the discrete event trigger mechanism. The main passive hybrid robust fault-tolerant control problem of linear uncertain NCS is studied by trigger mechanism. A NCS fault model based on discrete event triggered communication mechanism is established for linear uncertain NCS with time-varying delay. A robust fault detection observer based on H_ infinity control is designed, and a passive robust fault-tolerant controller is designed off-line to ensure the system. The system is stable when a known fault occurs, and the performance degradation speed of the system is slowed down at the beginning of the unknown fault, and the robust fault detection observer is used to detect the fault on line on line, and the effect of the controller to compensate for the effect of any unknown fault on the system is reconstructed.3) based on the event triggering mechanism, the robust H_ infinity control of the linear uncertain NCS primary and passive switching is studied. The system is a linear uncertain NCS with time-varying delay. Under the influence of the external finite energy disturbance, based on the discrete event triggered communication mechanism, the normal controller and the passive robust H_ infinity fault-tolerant controller are designed respectively, so that the system can have good dynamic performance when the system runs normally. When the system fails, the transient switching is passed. The function smooth switching to passive robust H_ infinity fault-tolerant controller ensures that the system is not only stable but also has a certain H_ infinity disturbance rejection performance when the known fault occurs, and the performance degradation speed of the system is slowed down at the beginning of the unknown fault, and the robust H_ infinity detection observer is designed, the fault is detected in real time, and the adaptive compensation control is used to eliminate the fault. The influence of the unknown fault on the system.4) on the basis of the above research, the simulation example is used to simulate the results of the passive fault-tolerant controller, the active fault-tolerant controller, the hybrid fault-tolerant controller and the switching function design, and the analysis of the influence of the system performance. The results show that the conclusions obtained in this paper are correct. For the above research conclusions, the simulation results show that the event generator is introduced at the sensor end and the design of fault estimation with time-varying delay linear uncertainty NCS and the integrated robust fault-tolerant control integration design with the primary and passive hybrid are realized. The proposed fault diagnosis method based on the observer can be proposed. It is effective to diagnose any fault. In different cases, the state feedback control strategy is adopted to design the passive hybrid robust fault-tolerant controller, which makes the system not only asymptotically stable but also asymptotically stable in both normal running state and actuator fault.
【学位授予单位】：兰州理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP273

【参考文献】