[ { "id": "authors:z1wwm-v7t37", "collection": "authors", "collection_id": "z1wwm-v7t37", "cite_using_url": "https://resolver.caltech.edu/CaltechAUTHORS:20180125-151615858", "type": "book_section", "title": "Thinking Fast and Slow: Optimization Decomposition Across Timescales", "book_title": "2017 IEEE 56th Annual Conference on Decision and Control (CDC)", "author": [ { "family_name": "Goel", "given_name": "Gautam", "orcid": "0000-0002-7054-7218", "clpid": "Goel-Gautam" }, { "family_name": "Chen", "given_name": "Niangjun", "orcid": "0000-0002-2289-9737", "clpid": "Chen-Niangjun" }, { "family_name": "Wierman", "given_name": "Adam", "orcid": "0000-0002-5923-0199", "clpid": "Wierman-A" } ], "abstract": "Many real-world control systems, such as the smart grid and human sensorimotor control systems, have decentralized components that react quickly using local information and centralized components that react slowly using a more global view. This paper seeks to provide a theoretical framework for how to design controllers that are decomposed across timescales in this way. The framework is analogous to how the network utility maximization framework uses optimization decomposition to distribute a global control problem across independent controllers, each of which solves a local problem; except our goal is to decompose a global problem temporally, extracting a timescale separation. Our results highlight that decomposition of a multi-timescale controller into a fast timescale, reactive controller and a slow timescale, predictive controller can be near-optimal in a strong sense. In particular, we exhibit such a design, named Multi-timescale Reflexive Predictive Control (MRPC), which maintains a pertimestep cost within a constant factor of the offline optimal in an adversarial setting.", "doi": "10.1109/CDC.2017.8263834", "isbn": "978-1-5090-2874-0", "publisher": "IEEE", "place_of_publication": "Piscataway, NJ", "publication_date": "2017-12", "pages": "1291-1298" }, { "id": "authors:1gjze-jr229", "collection": "authors", "collection_id": "1gjze-jr229", "cite_using_url": "https://resolver.caltech.edu/CaltechAUTHORS:20170515-140123828", "type": "book_section", "title": "Smoothed Least-laxity-first Algorithm for EV Charging", "book_title": "e-Energy '17 Proceedings of the Eighth International Conference on Future Energy Systems", "author": [ { "family_name": "Nakahira", "given_name": "Yorie", "orcid": "0000-0003-3324-4602", "clpid": "Nakahira-Yorie" }, { "family_name": "Chen", "given_name": "Niangjun", "orcid": "0000-0002-2289-9737", "clpid": "Chen-Niangjun" }, { "family_name": "Chen", "given_name": "Lijun", "clpid": "Chen-Lijun" }, { "family_name": "Low", "given_name": "Steven H.", "orcid": "0000-0001-6476-3048", "clpid": "Low-S-H" } ], "abstract": "We formulate EV charging as a feasibility problem that meets all EVs' energy demands before departure under charging rate constraints and total power constraint. We propose an online algorithm, the smoothed least-laxity-first (sLLF) algorithm, that decides on the current charging rates based on only the information up to the current time. We characterize the performance of the sLLF algorithm analytically and numerically. Numerical experiments with real-world data show that it has significantly higher rate of generating feasible EV charging than several other common EV charging algorithms.", "doi": "10.1145/3077839.3077864", "isbn": "978-1-4503-5036-5", "publisher": "ACM", "place_of_publication": "New York, NY", "publication_date": "2017-05", "pages": "242-251" }, { "id": "authors:akf6m-05e57", "collection": "authors", "collection_id": "akf6m-05e57", "cite_using_url": "https://resolver.caltech.edu/CaltechAUTHORS:20170110-153001767", "type": "book_section", "title": "Using Predictions in Online Optimization: Looking Forward with an Eye on the Past", "book_title": "SIGMETRICS '16 Proceedings of the 2016 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Science", "author": [ { "family_name": "Chen", "given_name": "Niangjun", "orcid": "0000-0002-2289-9737", "clpid": "Chen-Niangjun" }, { "family_name": "Comden", "given_name": "Joshua", "clpid": "Comden-J" }, { "family_name": "Liu", "given_name": "Zhenhua", "clpid": "Liu-Zhenhua" }, { "family_name": "Gandhi", "given_name": "Anshul", "clpid": "Gandhi-A" }, { "family_name": "Wierman", "given_name": "Adam", "clpid": "Wierman-A" } ], "abstract": "We consider online convex optimization (OCO) problems with switching costs and noisy predictions. While the design of online algorithms for OCO problems has received considerable attention, the design of algorithms in the context of noisy predictions is largely open. To this point, two promising algorithms have been proposed: Receding Horizon Control (RHC) and Averaging Fixed Horizon Control (AFHC). The comparison of these policies is largely open. AFHC has been shown to provide better worst-case performance, while RHC outperforms AFHC in many realistic settings. In this paper, we introduce a new class of policies, Committed Horizon Control (CHC), that generalizes both RHC and AFHC. We provide average-case analysis and concentration results for CHC policies, yielding the first analysis of RHC for OCO problems with noisy predictions. Further, we provide explicit results characterizing the optimal CHC policy as a function of properties of the prediction noise, e.g., variance and correlation structure. Our results provide a characterization of when AFHC outperforms RHC and vice versa, as well as when other CHC policies outperform both RHC and AFHC.", "doi": "10.1145/2896377.2901464", "isbn": "978-1-4503-4266-7", "publisher": "ACM", "place_of_publication": "New York, NY", "publication_date": "2016-06", "pages": "193-206" }, { "id": "authors:yg990-t9x53", "collection": "authors", "collection_id": "yg990-t9x53", "cite_using_url": "https://resolver.caltech.edu/CaltechAUTHORS:20150715-103043881", "type": "book_section", "title": "Online Convex Optimization Using Predictions", "book_title": "Proceedings of the 2015 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems", "author": [ { "family_name": "Chen", "given_name": "Niangjun", "orcid": "0000-0002-2289-9737", "clpid": "Chen-Niangjun" }, { "family_name": "Agarwal", "given_name": "Anish", "clpid": "Agarwal-A" }, { "family_name": "Wierman", "given_name": "Adam", "clpid": "Wierman-A" }, { "family_name": "Barman", "given_name": "Siddharth", "clpid": "Barman-S" }, { "family_name": "Andrew", "given_name": "Lachlan L. H.", "clpid": "Andrew-L-L-H" } ], "abstract": "Making use of predictions is a crucial, but under-explored, area of online algorithms. This paper studies a class of online optimization problems where we have external noisy predictions available. We propose a stochastic prediction error model that generalizes prior models in the learning and stochastic control communities, incorporates correlation among prediction errors, and captures the fact that predictions improve as time passes. We prove that achieving sublinear regret and constant competitive ratio for online algorithms requires the use of an unbounded prediction window in adversarial settings, but that under more realistic stochastic prediction error models it is possible to use Averaging Fixed Horizon Control (AFHC) to simultaneously achieve sublinear regret and constant competitive ratio in expectation using only a constant-sized prediction window. Furthermore, we show that the performance of AFHC is tightly concentrated around its mean.", "doi": "10.1145/2745844.2745854", "isbn": "978-1-4503-3486-0", "publisher": "Association for Computing Machinery", "place_of_publication": "New York, NY", "publication_date": "2015-06", "pages": "191-204" }, { "id": "authors:10c4g-rc417", "collection": "authors", "collection_id": "10c4g-rc417", "cite_using_url": "https://resolver.caltech.edu/CaltechAUTHORS:20170810-105540217", "type": "book_section", "title": "Distributional analysis for model predictive deferrable load control", "book_title": "53rd IEEE Conference on Decision and Control", "author": [ { "family_name": "Chen", "given_name": "Niangjun", "orcid": "0000-0002-2289-9737", "clpid": "Chen-Niangjun" }, { "family_name": "Gan", "given_name": "Lingwen", "clpid": "Gan-Lingwen" }, { "family_name": "Low", "given_name": "Steven H.", "orcid": "0000-0001-6476-3048", "clpid": "Low-S-H" }, { "family_name": "Wierman", "given_name": "Adam", "clpid": "Wierman-A" } ], "abstract": "Deferrable load control is essential for handling the uncertainties associated with the increasing penetration of renewable generation. Model predictive control has emerged as an effective approach for deferrable load control, and has received considerable attention. Though the average-case performance of model predictive deferrable load control has been analyzed in prior works, the distribution of the performance has been elusive. In this paper, we prove strong concentration results on the load variation obtained by model predictive deferrable load control. These results highlight that the typical performance of model predictive deferrable load control is tightly concentrated around the average-case performance.", "doi": "10.1109/CDC.2014.7040398", "isbn": "978-1-4799-7746-8", "publisher": "IEEE", "place_of_publication": "Piscataway, NJ", "publication_date": "2014-12", "pages": "6433-6438" }, { "id": "authors:ncfw3-zg456", "collection": "authors", "collection_id": "ncfw3-zg456", "cite_using_url": "https://resolver.caltech.edu/CaltechAUTHORS:20130904-124117730", "type": "book_section", "title": "Data Center Demand Response: Avoiding the Coincident\n Peak via Workload Shifting and Local Generation", "book_title": "SIGMETRICS '13 Proceedings of the ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems", "author": [ { "family_name": "Liu", "given_name": "Zhenhua", "clpid": "Liu-Zhenhua" }, { "family_name": "Wierman", "given_name": "Adam", "clpid": "Wierman-A" }, { "family_name": "Cheng", "given_name": "Yuan", "clpid": "Cheng-Yuan" }, { "family_name": "Razon", "given_name": "Benjamin", "clpid": "Razon-B" }, { "family_name": "Chen", "given_name": "Niangjun", "orcid": "0000-0002-2289-9737", "clpid": "Chen-Niangjun" } ], "abstract": "Demand response is a crucial aspect of the future smart grid.\nIt has the potential to provide significant peak demand reduction and to ease the incorporation of renewable energy\ninto the grid. Data centers' participation in demand response is becoming increasingly important given the high\nand increasing energy consumption and the flexibility in demand management in data centers compared to conventional\nindustrial facilities. In this extended abstract we briefly describe recent work in [1] on two demand response schemes to\nreduce a data center's peak loads and energy expenditure:\nworkload shifting and the use of local power generations. In\n[1], we conduct a detailed characterization study of coincident peak data over two decades from Fort Collins Utilities,\nColorado and then develop two algorithms for data centers\nby combining workload scheduling and local power generation to avoid the coincident peak and reduce the energy expenditure. The first algorithm optimizes the expected cost\nand the second one provides a good worst-case guarantee for\nany coincident peak pattern. We evaluate these algorithms\nvia numerical simulations based on real world traces from\nproduction systems. The results show that using workload\nshifting in combination with local generation can provide significant cost savings (up to 40% in the Fort Collins Utilities'\ncase) compared to either alone.", "doi": "10.1145/2465529.2465740", "isbn": "978-1-4503-1900-3", "publisher": "ACM", "place_of_publication": "New York, NY", "publication_date": "2013-06", "pages": "341-342" }, { "id": "authors:m4gqe-7j072", "collection": "authors", "collection_id": "m4gqe-7j072", "cite_using_url": "https://resolver.caltech.edu/CaltechAUTHORS:20161025-161438752", "type": "book_section", "title": "Real-time deferrable load control: handling the uncertainties of renewable generation", "book_title": "e-Energy '13 Proceedings of the fourth international conference on Future energy systems", "author": [ { "family_name": "Gan", "given_name": "Lingwen", "clpid": "Gan-Lingwen" }, { "family_name": "Wierman", "given_name": "Adam", "clpid": "Wierman-A" }, { "family_name": "Topcu", "given_name": "Ufuk", "clpid": "Topcu-U" }, { "family_name": "Chen", "given_name": "Niangjun", "orcid": "0000-0002-2289-9737", "clpid": "Chen-Niangjun" }, { "family_name": "Low", "given_name": "Steven H.", "orcid": "0000-0001-6476-3048", "clpid": "Low-S-H" } ], "contributor": [ { "family_name": "Culler", "given_name": "David", "clpid": "Culler-D" }, { "family_name": "Rosenberg", "given_name": "Catherine", "clpid": "Rosenberg-C" } ], "abstract": "Real-time demand response is essential for handling the uncertainties of renewable generation. Traditionally, demand response has been focused on large industrial and commercial loads, however it is expected that a large number of small residential loads such as air conditioners, dish washers, and electric vehicles will also participate in the coming years. The electricity consumption of these smaller loads, which we call deferrable loads, can be shifted over time, and thus be used (in aggregate) to compensate for the random fluctuations in renewable generation. In this paper, we propose a real-time distributed deferrable load control algorithm to reduce the variance of aggregate load (load minus renewable generation) by shifting the power consumption of deferrable loads to periods with high renewable generation. At every time step, the algorithm minimizes the expected variance to go with updated predictions. We prove that suboptimality of the algorithm vanishes as time horizon expands. Further, we evaluate the algorithm via trace-based simulations.", "doi": "10.1145/2487166.2487179", "isbn": "978-1-4503-2052-8", "publisher": "ACM", "place_of_publication": "New York, NY", "publication_date": "2013-05", "pages": "113-124" }, { "id": "authors:m2gmy-t0323", "collection": "authors", "collection_id": "m2gmy-t0323", "cite_using_url": "https://resolver.caltech.edu/CaltechAUTHORS:20130826-144607036", "type": "book_section", "title": "Optimal charging of electric vehicles in smart grid: Characterization and valley-filling algorithms", "book_title": "IEEE Third International Conference on Smart Grid Communications", "author": [ { "family_name": "Chen", "given_name": "Niangjun", "orcid": "0000-0002-2289-9737", "clpid": "Chen-Niangjun" }, { "family_name": "Quek", "given_name": "Tony Q. S.", "clpid": "Quek-Tony-Q-S" }, { "family_name": "Tan", "given_name": "Chee Wei", "clpid": "Tan-Chee-Wei" } ], "abstract": "Electric vehicles (EVs) offer an attractive long-term solution to reduce the dependence on fossil fuel and greenhouse gas emission. However, a fleet of EVs with different EV battery charging rate constraints, that is distributed across a smart power grid network requires a coordinated charging schedule to minimize the power generation and EV charging costs. In this paper, we study a joint optimal power flow (OPF) and EV charging problem that augments the OPF problem with charging EVs over time. While the OPF problem is generally nonconvex and nonsmooth, it is shown recently that the OPF problem can be solved optimally for most practical power grid networks using its convex dual problem. Building on this strong duality result, we study a nested optimization approach to decompose the joint OPF and EV charging problem. We characterize the optimal offline EV charging schedule to be a valley-filling profile, which allows us to develop an optimal offline algorithm with computational complexity that is significantly lower than centralized interior point solvers. Furthermore, we propose a decentralized online algorithm that dynamically tracks the valley-filling profile. Our algorithms are evaluated on the IEEE 14 bus system, and the simulations show that the online algorithm performs almost near optimality (< 1% relative difference from the offline optimal solution) under different settings.", "doi": "10.1109/SmartGridComm.2012.6485952", "isbn": "978-1-4673-0910-3", "publisher": "IEEE", "place_of_publication": "Piscataway, NJ", "publication_date": "2012-11", "pages": "13-18" } ]