Research opportunities arising from control and optimization of smart buildings

Qianchuan ZHAO

Department of Automation and TNList,Tsinghua University,Beijing 100084,China

Research opportunities arising from control and optimization of smart buildings

Qianchuan ZHAO

Department of Automation and TNList,Tsinghua University,Beijing 100084,China

Control and optimization plays an important role in automation science and engineering for buildings.The area is important due to the growing threat of global warming effects to the sustainable development of human society as a whole indicated by the Paris Climate Treaty and the fact the buildings account for about 40%of total energy consumption in US,for example[1].

Effective control and maintenance of Heating,ventilation and air conditioning(HVAC),elevators,sensing devices,and other assets in building systems is extremely challenging since they are running under uncertain environments(such as continuously changing weather,human behaviors,and random contingencies).Uncertainty increases the dimensions of control strategies or maintenance polices since time invariant ones cannot adapt a dynamic environment.Uncertainty also increases the time to evaluate performances and the feasibility of control strategies or maintenance polices.

Traditional way to address the problem is to use heuristics or simulation based solutions that works only for a small number of pre-determined scenarios and rely experienced human beings to make decision for unexpected scenarios.For example,fixed set-point for HVAC control,fixed parking schedule for elevators,fixed communication frequency assignment for wireless sensors,scheduled maintenance for assets.This usually leads to poor energy efficiency,low customer satisfactions,low communication spectrum utilization,high maintenance cost or out of service of assets under unexpected contingency.

There is an increasing demand to develop effective low complexity solutions that make the building control and facility maintenance problem at practice size tractable in real-time.

Internet of Things(IoT)[2]seems provide a promising technique solution to gather a lot of real-time information about building operation and relevant to building management.This provides a huge opportunity for improving building operation performance and reducing maintenance cost.

Below we discuss several recent progresses in here in Tsinghua University just to give a sense of how this could be done.This by no means a complete list of possible benefits provided by IoT.Part of the work is supported by our industrial partners.

We introduced a convex 2-D personalized thermal comfort model that can be learned from occupants’complaint feedback[3].New HVAC control strategies employing this model have demonstrated significant energy saving potential while keeping occupants’satisfaction in both signal and group occupant scenarios[4].The model has also been used in developing machine learning based integrated control strategies for cooling,lighting,and shading that beats traditional integrated control strategies[5].

We established lower bounds of group elevator dispatching performance with advance information with a series of single car dispatching problem and surrogate sub-gradient method(SSG)or its extension penalty surrogate subgradient method(PSS)[6].This approach of obtaining upper and lower bounds avoids the need to compute the optimal solution to the single car dispatching problem and improve the efficiency of elevator group dispatching algorithm evaluation and has helped United Technology Research Center in design processes of elevators.We also collaborate with Toshiba Corp.in developing rule-based energy saving group control algorithms for elevators[7].In 2015,Toshiba Corp.has started to sell elevator products with the algorithms developed by our team.

In smart buildings,data collection based on Internet of Things(IoT)introduces more and more wireless sensors.Wireless communication frequency spectrum has become a scare resource and demands an efficient spectrum sharing among wireless sensors and keeping acceptably low level of interference to licensed users.To do so,we developed opportunistic spectrum access strategies based on periodic channel sensing for secondary communication users(sensors for building application)[8].The theoretical advantage of the strategies is that it avoids the intractable partial observable Markov decision process(POMDP)formulation and admit analytical solution for optimal solutions under tight collision constraints[9].

Besides the above listed work on control strategy design,we also made progresses in condition based asset(e.g.,HVAC chillers and gas turbines)maintenance policies.Joint replacement policies can cut down cost but usually difficult to find due to large search space.We found a monotone property of replacement cost for ranked remaining life times[10].This narrow down the search of optimal strategy within a linear size subset of the entire exponential size space.The algorithms developed by the nominee’s team has led to successful implementation in the management of re-manufacturing service network of Pratt&Whitney[11].

In fact,the above just list a small number of possible cases where new fascinating control and optimization problems arise from the study of smart buildings.There are many more need exploration.

For example,the accurate occupant counting or localization problem is still a hot research topic[12].Information fusion by combining video,CO2,and other data source is natural framework.How to achieve affordable yet accurate and reliable occupant information is fundamental for many other building control and management applications and is still open.

Another challenging problem is how to explore the valuable pieces of information hidden in different subsystems in buildings and efficiently use them to pinpoint a complete picture of the whole building and generate an intelligent decision as a response to dynamic environments in real-time[13,14].

It is exciting to see more and more control and system scientists and engineers have started to do research in this area.One indication is that major worldwide academic organizations have introduced dedicated technical committees(TCs)to promote the research.For example,IEEE Robotics and Automation Society(RAS)has started a TC on smart buildings.Please visit http://www.ieee-ras.org/smart-building for more details.IFAC also started recently a TC on control for smart cities[15].

[1]Q.Jia,Q.Zhao,H.Darabi,et al.Smart building technology.IEEE Robotics&Automation Magazine,2014,21(2):18–20.

[2]J.Gubbi,R.Buyya,S.Marusic,et al.Internet of Things(IoT):A vision,architectural elements,and future directions.Future Generation Computer Systems,2013,29(7):1645–1660.

[3]Q.Zhao,Y.Zhao,F.Wang,et al.Preliminary study of learning individual thermal complaint behavior using one-class classifier for indoor environment control.Building and Environment,2014,72:201–211.

[4]Q.Zhao,Z.Cheng,F.Wang,et al.Experimental study of group thermal comfort model.IEEE International Conference on Automation Science and Engineering,Taiwan:IEEE,2014:1075–1078.

[5]Z.Cheng,Q.Zhao,F.Wang,et al.Satisfaction based Q-learning for integrated lighting and blind control.Energy and Buildings,2016,127:43–55.

[6]J.Sun,Q.Zhao,P.Luh,et al.Estimation of optimal elevator scheduling performance.IEEE International Conference ON Robotics and Automation,Orlando:IEEE,2006:1078–1083.

[7]J.Wang,Y.Shen,S.Wang,et al.Energy-saving algorithm for elevator group control system with cameras.Proceedings of the 11th World Congress on Intelligent Control and Automation,Shenyang:IEEE,2015:2654–2658.

[8]Q.Zhao,S.Geirhofer,L.Tong,et al.Opportunistic spectrum access via periodic channel sensing.IEEE Transactions on Signal Processing,2008,56(2):785–796.

[9]X.Li,Q.Zhao,X.Guan,et al.Optimal cognitive access of Markovian channels under tight collision constraints.IEEE Journal on Selected Areas in Communications,2011,29(4):746–756.

[10]L.Xia,Q.Zhao,Q.Jia.A structure property of optimal policies for maintenance problems with safety-critical components.IEEE Transactions on Automation Science and Engineering,2008,5(3):519–531.

[11]T.Sun,Q.Zhao,P.Luh,et al.Optimization of joint replacement policies for multipart systems by a rollout framework.IEEE Transactions on Automation Science and Engineering,2008,5(4):609–619.

[12]T.Labeodan,W.Zeiler,G.Boxem,et al.Occupancy measurement in commercial office buildings for demand-driven control applications:A survey and detection system evaluation.Energy and Buildings,2015,93:303–314.

[13]F.Xiao,C.Fan.Data mining in building automation systems for improving building operational performance.Energy and Buildings,2014,75:109–118.

[14]F.Cheng,X.Fu,C.Yan.A framework for knowledge discovery in massive building automation data and its application in building diagnostics.Automation in Construction,2015,50:81–90.

[15]IFAC.Control for Smart Cities.2016:http://tc.ifac-control.org/9/3.

E-mail:zhaoqc@tsinghua.edu.cn.

This work was supported by the National Key Research and Development Project of China(No.2016YFB0901901)and the National Natural Science Foundation of China(No.61425027).

?2017 South China University of Technology,Academy of Mathematics and Systems Science,CAS,and Springer-Verlag Berlin Heidelberg

the B.E.degree in Automatic Control in July 1992,the B.S.degree in Applied Mathematics in July 1992,and M.Sc.and Ph.D.degrees in Control Theory and its Applications in July 1996,all from Tsinghua University,Beijing,China.He is currently a Professor and Director of the Center for Intelligent and Networked Systems(CFINS),Department of Automation,Tsinghua University.He was a Visiting Scholar at Carnegie Mellon University,Pittsburgh,PA,and Harvard University,Cambridge,MA,in 2000 and 2002,respectively.He was a Visiting Professor at Cornell University,Ithaca,NY,in 2006.His current research focuses on control and optimization of complex networked systems with applications in smart buildings,smart grid and manufacturing automation.He has published more than 80 research papers in peer-reviewed journals and conferences.He

the 2009 China National Nature Science Award for the project “Optimization Theory and Optimization for Discrete Event Dynamic System”.Dr.Zhao is a deputy Editor-in-Chief of Chinese version of the journal Control Theory&Applications,an associate editor for Journal of Optimization Theory and Applications,for IEEE Transactions on Control of Network Systems,and for IEEE Transactions on Automation Science and Engineering.He serves as a chair of the technical committee on smart buildings of IEEE RAS.E-mail:zhaoqc@tsinghua.edu.cn.