深圳大学管理学院20周年院庆系列学术讲座(第29期)暨深圳大学管理科学前沿论坛(第45期)

来源: 发布时间:2017-09-01 14:47:26 浏览量:
讲座题目:A Data-driven Adaptive Decision Framework for Distributed Building Clusters Online Operation
主讲人:胡孟琦,助理教授,伊利诺伊大学芝加哥分校
讲座时间:2017年9月3日(周日),15:00-16:00。
讲座地点:文科楼管理学院资料室
主持人:楚湘华,讲师

Abstract

According to the Electric Power Research Institute, electricity consumption in the U.S. has increased 1.7% annually from 1996 to 2006, with an expectation of total growth through 2030 to be 26%, with buildings responsible for over 70% of all electricity consumption. The emerging technologies in smart building, smart grid, renewable energy, as well as distributed energy resources drive research moving from centralized operation decisions on a single building to decentralized decisions on temporally and spatially distributed building clusters which share energy resources locally and globally. Optimizing energy systems operation in the building clusters will result in cost effective buildings which in turn will reduce overall primary energy consumption and peak time electricity consumption, and make buildings more resilient to power disruptions.
 
In this research, we first develop a mixed integer non-linear multi-objective decision model based on a building clusters agent-based simulator with each building modeled by energy consumption, storage and generation sub modules. Secondly, we propose a bi-level distributed decision framework based on an augmented multi-objective particle swarm optimization (AMOPSO) to study the tradeoff in energy usage among the group of buildings. AMOPSO is augmented via the fusion of multiple search methods to enhance PSO’s search capability on a diverse set of search spaces. To enable the online operation decision for building clusters, we extend the AMOPSO based decision approach to develop a data-driven adaptive decision framework for building clusters operation, through the use of noise-tolerant data fusion techniques to integrate real-time data collected from sensors and smart meters. The adaptive decision framework is demonstrated to be capable of deriving Pareto solutions for building clusters in a distributed manner. The operation decisions could not only reduce energy cost, but also improve environmental sustainability and buildings’ resilience capability to power disturbances.
 

Bio

Dr. Mengqi Hu received his Ph.D. and M.S. degrees in Industrial Engineering from Arizona State University in 2012 and 2010 respectively. He has also received his M.S. and B.S. degrees in Material Science and Engineering from Huazhong University of Science and Technology, China, in 2008 and 2006, respectively. He joined the Department of Mechanical and Industrial Engineering at the University of Illinois at Chicago (UIC) as a tenure-track Assistant Professor in August 2015. Prior to working at UIC, he has worked as a tenure-track Assistant Professor from August 2012 to August 2015.
 
His main areas of interests are complex system design and optimization, distributed decision support, swarm intelligence and system analytics, with applications to energy, manufacturing and healthcare systems. Most of his research projects are funded by NSF, DOD, DOT, and NSA. He is a recipient of 2015 AFOSR summer faculty fellow award. He has published 22 high quality journal articles and 10 peer-reviewed conference papers.
 
 欢迎感兴趣的师生参加!
 
深圳大学管理学院管理科学系
2017年9月1日
电话:+86-755-26536121 传真:+86-755-26534451
Copyright © 2003-2015 All rights reserved
官方微信号