具有通信约束的网络化关联大系统分散控制及协同演化分析

It is well known that communication constraint is one of the bottleneck problems in restricting further development of networked systems. Networked system have been received much attention in the fields of complex large-scale system control, remote control and so on due to the rapid development and integration of network, communication and control technology. A lot of new problems have arisen such as the information congestion, packet loss, network- induced delay, quantization error, etc due to communication constraints and limited channel capacity of the network, which are often one of the main causes of instability and poor performance of the systems. The traditional control theories and technologies have become increasingly unsuitable for these new characteristics of the scenarios. Therefore, it is an urgent requirement currently to study new models, theories and methods to solve the problems of decentralized control and cooperative control for interconnected large-scale networked control systems with communication constraints. To this end, this project is mainly targeted to investigate in the following aspects: 1) Some new system models are proposed for interconnected large-scale networked control systems with communication constraints by analyzing of the mechanism characteristics of the critical factors in the network environment; 2) Based on the proposed models, some novel effective stability criteria and practical decentralized and cooperative control algorithms are given by through strict dynamic analysis; 3) Simulations on study of networked smart micro grids large-scale system under a specific network communication environment will be demonstrated to improve the proposed algorithms of the project. This project has a very important place in accelerating the development of next-generation communications networks and control technology, promoting the study of information theory, and expediting the original innovation of the network technologies.

通信约束是制约网络化系统进一步发展的瓶颈问题之一。网络化大系统因网络、通信与控制技术的迅猛发展与结合，在复杂大系统控制、远程控制等领域广受关注。受通信约束及信道容量有限等因素影响，带来了信息拥塞、丢包、网络时滞、量化误差等新问题，它们会导致系统性能下降甚至不稳定。传统的理论技术已越来越不适应新的环境特点，研究新的模型、理论与方法来解决受限通信环境下网络关联大系统的分散协同控制问题已成为当前的迫切要求。为此，本项目主要从以下几方面进行针对性研究：1)通过分析网络环境下各种关键因素的机理特性，建立具有通信约束的网络化关联大系统模型；2)对上述模型进行严格的动力演化分析，建立有效稳定判据并提出实用的分散协同控制算法；3)通过特定通信环境下智能微电网大系统进行仿真研究，对现有算法提出改进方案。本项目对于加快新一代通信网络和控制技术的发展、推动信息理论的研究、促进通信网络技术的原始创新均有重要意义。

本项目针对通信约束和不完全信息的网络化关联大系统的分散控制和协同分析多项关键科学问题，在基础理论研究及实际仿真验证和应用等方面取得了一系列成果。发表标注项目资助的论文56篇，其中SCI收录27篇，EI收录51篇在IEEE Trans. 汇刊上发表长文4篇；入选上海曙光人才计划；获批国家自然基金面上项目1项及省部级课题3项。本项目结合通讯约束及不完善信息，对网络化关联系统的分散控制和滤波的分析和设计问题展开研究，同时为验证所提理论方法的有效性，对研究成果进行了仿真验证。不完全信息包括传感器饱和、数据包丢失、随机噪声以及量化误差等。研究基于不完全信息的情况下，对于一些网络控制系统模型，设计有效的控制器和滤波器是很有必要的。因为网络控制的效率高和灵活性好，现代控制越来越依赖NCSs。对于一个工业控制系统来说，发生故障是不可避免的。由此，学者们提出了一系列的基于网络的故障诊断方法。在不完全信息的条件下，NCSs的故障诊断更为复杂，要求更高。因此，本项目研究了在基于不完全信息的情况下NCSs的H∞控制、滤波与故障诊断问题。. 项目主要成果包括：1）针对具有混合时滞(分布式时滞、状态时滞)与量化误差的网络控制系统模型，利用Lyapunov稳定性理论与线性矩阵不等式(LMI)的方法设计一个H∞控制器；2）研究了具有传感器饱和、随机发生的非线性与随机发生的分布式时滞的网络化系统，用满足Bernouli分布的白噪声来描述通道中的丢包现象。通过Lyapunov理论来得出系统稳定的充分条件，得到系统基于LMI的可行解，设计出一个H∞滤波器；3）研究了一类具有乘性噪声、传感器饱和与分布式时滞的网络化系统的故障诊断问题。通过引入故障评估函数，并将此性能指标与评估阈值函数比较，完成滤波系统的故障诊断。针对不完全信息的故障诊断，研究并设计出满足H∞性能水平的滤波器。4）基于不完全信息的NCSs控制、滤波和故障诊断的理论研究成果在网络化系统中的应用和仿真验证。