基于低复杂度预测控制的网络控制系统理论与应用研究

Flexibility, reconfigurability, low installation cost and ease of maintenances are some of the advantages of networked control systems (NCS), which have been recognized as a very challenging and promising research field. While possessing some well-known advantages, the architectures of NCS can give rise to several common control issues, such as network delay, data dropout, and mis-synchronization etc. To solve these problems, various methods have been developed. In order to model the time delay and data loss accurately, however, these methods have put some strict assumptions on NCSs. Most of these methods ignores NCS feature and cannot compensate for the time delay and data dropout. One interesting feature of modern communication is that data is sent in large packets. In this context, it makes sense to send signal predictions, which are calculated from model predictive control (MPC) method at the transmission side, to compensate for time delays and data-dropouts. The big drawback of the MPC is the formidable on-line computational effort, which limits its applicability to relatively slow dynamics. In recent years, the low complexity MPC is developed, which can reduce the on-line computational burden. But the developed complexity MPC concerns with linear systems, which cannot directly apply to NCS. .This project is concerned with networked time delay,data dropout and input saturation. Applying the low complexity MPC is expected to work out the problems mentioned above. We have reason to believe that this project will be of great value to theory and applications.

网络控制系统（NCS）具有灵活性和可扩展性，低安装成本，易于维护等优点，是一个非常具有前景和挑战性的研究领域。尽管有许多优点，通信网络的引入带来很多新的问题。例如，NCS中存在网络时滞、数据包丢失、时间不同步等等。很多方法在处理这些问题的时候，为了对网络中的时滞与丢包进行精确建模，必须对其作一些比较严格的限制。这些方法，未能有效利用NCS数据传输的特点，无法对系统中存在的时滞和丢包进行补偿。利用网络中以包为单位进行数据传输的特点，以及模型预测控制（MPC）的预测特性可以有效的补偿系统的时滞和丢包。但是MPC的在线计算量大，仅仅适用于慢变系统。近年来发展起来的低复杂度MPC可以降低计算复杂度，减少在线计算时间，但是目前的算法均基于线性系统，不能直接应用于NCS中。.针对NCS，采用低复杂度MPC算法，对网络时滞、丢包以及输入饱和进行研究，有望解决上述问题。本项目具有重要的理论和应用价值。

随着计算机技术、通信技术与控制技术的发展，网络控制系统的研究具有重要的理论价值和广泛的应用前景。本项目以网络控制系统为研究对象，考虑了数据的采样、参数不确定性、饱和输入等物理与通信方面的约束，采用预测控制和鲁棒控制对其进行分析设计。旨在提高系统的稳定性、鲁棒性和控制性能。经过广泛学习和深入研究，相关研究成果弥补了该领域的空白。对网络控制系统的应用和发展，提供了理论依据与技术支持。