专用集成电路与系统国家重点实验室
讲座信息
题目：Control and Estimate of Uncertain Systems
Part 1. Flexible Robust Sliding Mode Control for Uncertain Stochastic Systems with Time-Varying Delay and Structural Uncertainties
Part 2. Novel Estimation Method of Community AADT and VMT via Circuit Network Models and Their Simulation
报告人：王胜国教授（美国北卡罗来纳大学）
时间：2013年5月15日（周三）上午10:00-11:00
地点：张江校区微电子楼369室
Abstract
Uncertainties and disturbance do commonly exist in various systems due to the changing environment and process variation as well as various causes, e.g., VLSI process variation in nano-size. In addition to that, the time-delay is common in remote control systems, communications, internet networks and computing systems, e.g., remote moon-vehicle. Questions in this talk we treat are: How to robustly control uncertain stochastic systems? How to estimate the networks performance without monitors and sensors?
Part 1. Q: How to robustly control uncertain time-delay stochastic systems?
A: Flexible Robust Sliding Mode Control for Uncertain Stochastic Systems with Time-Varying Delay and Structural Uncertainties
This talk considers very general uncertain stochastic systems with time-varying delay, structural uncertainties and the Brownian noise, and presents a new robust sliding mode control (SMC) method for this kind of general uncertain time-delay stochastic systems.
The proposed method applies singular value decomposition (SVD) to all the structural uncertainties, and introduces adjustable parameters for control design. A less-conservative condition of robust stability and new robust controller for the uncertain stochastic systems are developed via linear matrix inequality (LMI) forms. The reachability of system states to the SMC switching surface is guaranteed in probability of one by the proposed control rule. It is theoretically proved that the proposed method is less conservative than the existing methods in the literature. The simulation examples also demonstrate the correctness of the results.
Part 2. Q: How to estimate the networks performance, e.g., average transportation AADT in areas without monitoring system?
A: Novel Estimation Method of Community AADT and VMT via Circuit Network Models and Their Simulation
Transportation is essential and important to peoples in modern times. Annual Average Daily Traffic (AADT) and Vehicle Miles Traveled (VMT) of traffic networks are essential data for traffic plan and decision-making. Because of lack of traffic monitor system and the complicity of local area road networks, how to estimate the AADT and VMT on local roads is a national recognized long-time existing difficult problem. To solve this problem, this talk presents a novel AADT and VMT estimation approach via circuit networks models. In this approach, we convert traffic networks into circuit networks, and simulate the circuit networks for the AADT and VMT estimate of local community area traffic networks. Three sub-models based on three different distributions of the entrance traffic flows among households in a community are developed and combined together by the least squares method. The experiment results validate our approach. Furthermore, this method is able to estimate AADT and VMT on community area networks even without any traffic monitor, and meanwhile dramatically reduce the labor and equipment costs.
