无线网络化控制系统中能量与通信受限的多传感器协作估计研究

Wireless Networked Control Systems (WNCSs) has the low-cost, easy-to-deploy and easy-to-expend characteristics. It has become one of the hot topics in the control field. However, in the application of wireless technology to control systems, the energy and communication constraints will affect the control system's overall performance. The existing wireless networked control systems theory is not yet mature and cannot solve these problems. This proposal will devote to improving WNCSs’ state estimation with energy and communication constraints. It will start from studying the stochastic changes of the multi-sensor cooperative estimation's covariance. Using the stochastic processes and matrix theory, it will analyze the estimator's stability and find the condition when error covariance is bounded; using stochastic dynamic programming method, design periodic and dynamic cooperative sensor scheduling to optimize the estimation performance, to promote sensors’ collaboration; according to energy constraint, reduce the network energy consumption and improve energy utilization, under which the estimation performance is guaranteed. The results will be changed into actionable protocols for the network environment. The simulation platform and experimental platform will be utilized to valid and further perfect the theory and application framework. This proposal will expect to make a contribution to the WNCSs in the practical application in industrial automation.

无线网络化控制系统(WNCSs)具有低成本、易部署和易扩展等特点，目前已成为控制界的热点研究课题之一。然而，在将无线网络技术应用于控制系统时，传感器能量不足、网络通信受限等局限性都对控制系统整体性能产生了影响，现有的WNCSs研究还不能有效解决这些问题。为此，本项目将致力于提高WNCSs在能量与通信受限约束下的状态估计性能。从分析多传感器协作估计中估计误差协方差的随机变化出发，运用随机过程和矩阵论，分析估计器的稳定性，寻找保障误差有界的条件；运用随机动态优化方法设计多传感器周期的、动态的协作调度策略优化整体估计性能，促进多传感器相互协作；针对能量受限问题，提出在保证估计性能的前提下，降低网络能耗，提高能量利用率达到公平分配资源的目的。并将所得结论转化成能在网络环境中可实施的算法、协议，发挥仿真平台和实验平台的作用，进一步验证和完善理论与应用框架，推动WNCSs在工业自动化的实际应用。

在无线网络化控制系统中，控制系统整体性能受到传感器能量不足、网络通信受限等约束，本项目致力于提高WNCSs在能量与通信受限约束下的状态估计性能。分析多传感器协作估计中估计误差协方差的概率特征，运用随机过程和矩阵论，分析估计器的稳定性，保障误差有界的充分必要条件。设计多传感器周期、动态协作调度策略促进多传感器相互协作，优化整体估计性能。针对能量受限，提出在保证估计性能的前提下，降低网络能耗，提高能量利用率达到公平分配资源的目的。