网络化系统中采样周期问题的研究

发布时间：2019-01-05 18:43

【摘要】：网络化控制系统(networked control system,NCS)是集通信网络和控制系统于一体的复杂系统,它借助于通信网络将分布于不同地理位置的传感器、控制器和执行机构连接起来,形成一种全分布的实时反馈闭环控制系统。在NCS中,采样周期的选择决定了网络交换信息的频繁程度,当采样周期越小,数据传输的速率就越快,控制器得到的反馈信息就越多,从而控制品质(Quality of Performance,Qo P)就越好,但是,过快的数据传输会使得网络具有较高的吞吐率,导致网络拥塞,加重网络负载和时延使得网络服务质量(Quality of Service,Qo S)变差;相反,采样周期太长,虽然能够提高Qo S,但是不能保证实时数据的传送,会降低NCS的Qo P。所以,采样周期是连接NCS的Qo P和Qo S的纽带,为了改善NCS的性能必须选择合适的采样周期,本文针对NCS中的采样周期相关问题,主要做了以下工作:首先,从定常采样周期、时变采样周期和多速率采样三个方面整理了近年来NCS中采样周期的研究进展和成果以及目前关于NCS中采样周期的调度方法。其次,分析了NCS中时延的组成以及产生时延的主要因素,利用Lyapunov理论和LMI分别对基于连续模型和离散模型的NCS进行了时延相关稳定性分析,并且应用MATLAB进行了实例仿真验证。再次,在不考虑时延的情况下对连续系统的二次型性能函数离散化,得到不具有时延的离散二次型性能函数,并且在此基础上对同一被控对象具有时延的单采样率和多速率的二次型性能指标函数进行了详细推导,得到表示各指标函数中的权矩阵与连续二次型中权矩阵的直接关系。最后,根据具有时延的NCS的增广状态空间描述模型,利用第二章的时延相关稳定性定理假设系统的时延为最大允许时延,通过求解第三章研究中的二次型优化函数的极值,并且结合经验公式和稳定性条件得到采样周期的优化解。
[Abstract]:The networked control system (networked control system,NCS) is a complex system that integrates the communication network and the control system. By means of the communication network, it connects sensors, controllers and actuators distributed in different geographical locations. A fully distributed real-time feedback closed-loop control system is formed. In NCS, the selection of sampling period determines the frequency of network information exchange. The smaller the sampling period, the faster the rate of data transmission, the more feedback the controller gets, and the better the quality of (Quality of Performance,Qo P is. However, too fast data transmission will make the network have higher throughput, lead to network congestion, increase the network load and delay, make the network quality of service (Quality of Service,Qo S) worse; On the contrary, the sampling period is too long, although it can improve the Qo, but it can not guarantee the transmission of real-time data, it will reduce the Qo of NCS. Therefore, the sampling period is the link between Qo P and Qo S of NCS. In order to improve the performance of NCS, we must select the appropriate sampling period. In this paper, we mainly do the following work in order to solve the problems related to the sampling period in NCS: first, from the point of view of steady sampling period, The research progress and achievements of sampling period in NCS in recent years and the scheduling methods of sampling period in NCS are summarized in three aspects: time-varying sampling period and multi-rate sampling. Secondly, the composition and main factors of delay in NCS are analyzed. The time-dependent stability of NCS based on continuous model and discrete model is analyzed by using Lyapunov theory and LMI, respectively, and an example of MATLAB is used to verify the stability of the delay-dependent stability. Thirdly, the quadratic performance function of continuous system is discretized without considering the delay, and the discrete quadratic performance function without delay is obtained. On this basis, the quadratic performance index function with single sampling rate and multi-rate delay for the same controlled object is derived in detail, and the direct relationship between the weight matrix in each index function and the weight matrix in the continuous quadratic form is obtained. Finally, according to the augmented state space description model of NCS with delay, using the delay-dependent stability theorem in Chapter 2, the delay of the system is assumed to be the maximum allowable delay, and the extremum of the quadratic optimization function in Chapter 3 is solved. The optimal solution of sampling period is obtained by combining empirical formula and stability condition.
【学位授予单位】：兰州理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP273

【相似文献】