*主持人：李韬 教授

*讲座内容简介：
Optimization problems in network systems must resolve issues of computation and communication complexity, unknown system models, and observation noise. In this talk, we discuss different optimization structures and their implications on complexity, scalability, privacy, and security. We show that by employing common distributed features on physical systems and integrating cyber and physical systems, it is possible to derive strictly distributed cyber optimization algorithms that have highly desirable features. When physical system models are either unknown or change with time or operating conditions, this cyber-physical integration approach provides operational data that can be used to identify unknown systems during optimization, leading to adaptive strictly distributed optimization algorithms. However, the integrated algorithm of system identification (learning) and optimization encounters a fundamental conflict between system identifiability and convergence to optimal solutions. To maintain persistent identifiability and achieve optimality, we introduce a method of adding a small stochastic dither which provides sufficient excitation for estimating the unknown parameters, leading to convergent adaptive optimization algorithms. Joint identification and optimization algorithms are developed and their simultaneous convergence properties of parameter estimation and optimization variable updates are proved. Under both noise-free and noisy observations, the corresponding convergence rates are established. The main results of this paper reveal certain fundamental relationships and trade-off among updating step sizes, dither magnitudes, parameter estimation errors, optimization accuracy, and convergence rates. Simulation case studies are used to illustrate the adaptive optimization algorithms and their main properties.

*主讲人简介：
王乐一，1990获得加拿大麦吉尔大学博士学位。自1990年，他任职于美国韦恩州立大学，现为该校电气与计算机工程系教授。他的研究领域包括复杂性与信息，系统辨识，鲁棒控制，信息处理和学习，及其在医药、汽车、通信和电力系统中的应用。他曾在多个国际会议做大会报告，是国际自动控制联合会建模、辨识与信号处理专委会委员，曾担任包括IEEE Transactions on Automatic Control 在内的多个期刊的责任编委。王乐一教授是IEEE Fellow。