How Can Smart Buildings Be Price-Responsive?

Juan Miguel Morales; Ricardo Fern\'andez-Blanco; Salvador Pineda

REVIEW 3 major objections 4 minor 21 references

How Can Smart Buildings Be Price-Responsive?

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Widening the comfort bounds an occupant allows in a smart home shifts roughly 7–10 percentage points of annual electricity use into low-price periods and cuts annual electricity bills by 16–34%, the paper contends.

desk verdict A solid integrated economic MPC formulation for price-responsive households, with ideal-planning savings that need a forecast-error robustness test before the quantitative claims can be taken at face value. read the letter →

arxiv 1908.00481 v1 pith:F5R5UYJQ submitted 2019-08-01 eess.SY cs.SY

Ricardo Fernández-Blanco , Juan Miguel Morales , Salvador Pineda This is my paper · ORCID

classification eess.SYcs.SY

keywords modelpredictivecontrolprice-responsiveloadssmartbuildingsappliancescomfortconstraintsbuildingthermaldynamicsdemand-sideflexibilityeconomicMPC

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that the price responsiveness of a smart building is governed less by the heating hardware than by the comfort bounds the occupants allow the controller. It formulates an economic model predictive control problem for a single-zone household—heating or cooling, water heater, refrigerator, lighting, and uninterruptible appliances—and simulates a full year of 15-minute decisions using known day-ahead prices and one-day weather forecasts. In these simulations, widening the acceptable temperature range from zero to a few degrees raises the share of annual electricity used in low-price periods from 54.5–57.2% to 64.1–64.7% and lowers annual electricity cost by 16–34% depending on the heating system. The paper also finds that making the comfort bounds price-dependent reduces occupant discomfort relative to constant bounds, but at a higher electricity cost.

What carries the argument

The central object is an economic model predictive control (MPC) formulation: a mixed-integer linear program that, at each 15-minute step, minimizes electricity cost plus penalty terms for comfort violations over a one-day look-ahead horizon. The building is represented by a five-state linear thermal model—room air, floor, water in the floor-heating pipes, refrigerator chamber, and water-heater temperature—with external disturbances for ambient temperature, occupancy, and hot-water demand. Comfort enters as user-defined, time-varying upper and lower bounds on the temperatures and on light level, with slack variables penalized in the objective, so the controller is allowed to violate bounds at a price. Uninterruptible loads such as the washing machine and dishwasher are scheduled by binary variables with consecutive cycle constraints and phase-specific power draws. This machinery does the work of shifting consumption to low-price periods by preheating or precooling the building's thermal mass when electricity is cheap and deferring non-thermal loads.

What would settle it

Run the same economic MPC controller for a year in a real instrumented household (or a high-fidelity simulation with realistic forecast errors), alternating between tight and wide comfort bounds under the same tariff, and compare the annual share of electricity consumed in low-price periods. If the share does not increase by roughly 7–10 percentage points, or the annual cost does not fall by near 16–34%, the paper's central claim fails.

Watch

Extended reading notes

Core claim

In the paper's own terms, the central discovery is that occupant-defined comfort constraints are the key driver of a household's price-responsive behaviour. Using the reported three flexibility cases, the annual cost falls from €103.5 to €68.4 for a water-based floor heater (a 34% saving) and from €107.0 to €76.1 for an HVAC system (a 29% saving) when comfort bounds widen from no flexibility to extra flexibility under price-independent bounds. At the same time, the share of building consumption in low-price periods rises from 54.5–57.2% to 64.1–64.7% for both systems. A separate result is that a smaller heat-transfer coefficient (better insulation) makes HVAC-equipped households more price-responsive, whereas floor-heated households become more price-responsive when less insulated; the latter is attributed to the slow thermal dynamics of the floor mass.

Load-bearing premise

The load-bearing premise is that the controller's one-day look-ahead planning on a perfect five-state linear thermal model with known prices, weather, occupancy, and hot-water demand predicts how the real building and its occupants respond; if forecasts are imperfect or the model omits solar gain, humidity, and variable heat-pump efficiency, the simulated savings and price-shift will not be realized.

Editorial extensions

If this is right

A household can achieve substantial price responsiveness without adding batteries, solar panels, or any new equipment; the only change is widening the allowed comfort band in the controller.
The same MPC formulation handles thermostatic and non-thermostatic loads jointly, so dishwasher, dryer, and water-heater schedules shift together with heating and cooling to cheap periods.
Price-dependent comfort bounds offer a direct trade-off: occupants who care about staying near the reference temperature get less discomfort with PD-CB, but pay more than they would with uniformly wide, price-independent bounds.
The interaction between insulation and heating type matters: for fast HVAC systems, better insulation increases price responsiveness, while for slow water-based floor heating, worse insulation increases it, at higher annual cost.
Annual simulations with 15-minute decisions and integer appliance scheduling are computationally tractable (resolved in 16–60 minutes per case on one CPU), so the controller could be re-run daily in practice.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the simulated 7–10 percentage point shift in low-price consumption survives real forecast error, aggregating many such homes could flatten the daily price curve and reduce peak generation needs; this is an extension the paper does not test.
The perfect-forecast, linear-model idealization means the reported savings are likely an upper bound for field operation; a testable extension is an A/B field trial with real day-ahead prices and forecast updates.
The sensitivity to the room-ambient heat-transfer coefficient suggests that insulation level and heating-system choice should be co-optimized with comfort settings under a given tariff; this joint design problem is implicit in the results but not solved here.
A natural next step not explored in the paper is to compare this comfort-bound flexibility against battery storage on a cost-per-flexibility basis for the same household.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The useful thing here is the integrated model: five-state thermal dynamics for two heating systems plus refrigerator, water heater, lighting, and phase-dependent uninterruptible loads in one economic MPC, run over a full year. The discrete phase scheduling for appliances is a real addition to the building-MPC literature. The sensitivity analysis of comfort bands, price-dependent bounds, and insulation is also done carefully. On those terms the paper is a solid extension of Halvgaard and related work. Cost savings of 16–34% and a 6–10 point shift into low-price periods are plausible within the model.

The soft spots are real but not disqualifying. The building model is a linear grey-box taken from earlier papers, not validated against measured data. The MPC is run with perfect day-ahead forecasts of weather, occupancy, hot-water demand, and prices. That matters: the reported numbers are ideal-planning results, not closed-loop outcomes under uncertainty. A forecast-error test would tell us how much of the 16–34% survives. The low-price threshold is arbitrary, and the results come from one household, one weather year, and one price series, so the quantitative claims should be read as case-study results, not general bounds. I also note the main direction is built into the objective: widening comfort bounds has to lower cost in an economic MPC. What is not built in is the magnitude, and that is what the paper estimates.

The paper does not oversell much; the conclusions mostly stay close to the simulations. The auxiliary data link is good practice. The price-dependent comfort bounds result, where flexibility costs more but reduces discomfort, is a nice nuance.

Who is this for: people building residential demand-response models or needing a compact benchmark formulation. I would not treat the quantitative savings as empirical evidence. I would treat the model and sensitivity setup as a usable starting point.

Recommendation: serious referee, yes. The main revision request should be a forecast-error sensitivity study and an explicit caveat that the current numbers are perfect-information upper bounds. If the authors add that, the paper is publishable as a modeling and benchmark contribution.

Referee Report

3 major / 4 minor

Summary. The paper proposes a compact economic model predictive control (MPC) formulation for a single-zone smart household that includes a five-state thermal model (room, floor, floor-water, refrigerator, water heater), two alternative space-heating systems (water-based floor heater and HVAC), lighting, and interruptible appliances with discrete scheduling constraints. The authors simulate one year of operation under perfect day-ahead forecasts, comparing three flexibility levels (noﬂex, ﬂex, extraﬂex) and two comfort-bound strategies (price-independent and price-dependent), and report annual electricity cost, temperature-bound violations, and the share of consumption occurring in low-price periods. They conclude that widening comfort bounds produces cost savings of 16–34% depending on heating system and strategy, and shifts 6–10 percentage points more of annual consumption into low-price periods. They also analyze the effect of the building's heat-transfer coefficient on these outcomes.

Significance. If the reported magnitudes hold under realistic operating conditions, the paper would provide useful estimates of the demand-side flexibility potential of smart homes, a topic of current interest for distribution system operators and aggregators. The paper's strengths are its integrated formulation that combines thermal dynamics, appliance scheduling, and comfort constraints in a single optimization model; the explicit modeling of uninterruptible loads with variable power cycles; the comparison of two space-heating technologies; and the parametric study of insulation level. The authors also provide a data link for reproducibility. However, the quantitative claims rest on a perfect-forecast assumption, an unvalidated linear thermal model, and a single year of data, which limits the external validity of the specific percentages; the contribution is best viewed as an idealized upper-bound analysis rather than a robust prediction of achievable savings.

major comments (3)

[Section IV and Section V] The simulations use perfect day-ahead knowledge of ambient temperature, solar radiation, occupancy, hot-water demand, and electricity prices, as stated in Section IV ('we use a look-ahead window of one day' and the historical profiles of [18]). No sensitivity analysis with forecast error is provided. Since economic MPC under imperfect forecasts will preheat or shift load at suboptimal times, the central quantitative claims in Section V (cost savings of 16–34% and the increase in low-price-period consumption share from 54.5–57.2% to 64.1–64.7%) are idealized upper bounds. To support the practical framing, the authors should include robustness tests with, for example, additive noise on prices, temperature, and occupancy forecasts, or with a stochastic MPC formulation.
[Section III, Eq. (4), and Section IV] The five-state linear thermal model in Eq. (4) is adopted from [8] and [18] without validation against measured data for the simulated household. The only structural sensitivity study, in Section IV-B, varies the room-to-ambient heat transfer coefficient UAr,a while all other parameters and the model structure (constant COP, no humidity, no thermal solar-gain term) remain fixed. Because the reported savings and load-shift percentages are specific to this assumed model, the paper should either validate the model against data or clearly state that the results are illustrative of the model, and ideally include sensitivity analysis over other thermal parameters.
[Section IV-A (low-price-period definition)] The price-responsiveness metric is defined by an arbitrary threshold: low-price periods are those where the annual normalized price is lower than 0.5. The central claim in Section V that the share of building consumption in low-price periods rises from 54.5–57.2% to 64.1–64.7% depends on this threshold. No sensitivity to the threshold value is reported, so it is unclear whether the observed shift is robust or partly an artifact of the chosen cutoff. The authors should report the consumption share for a range of thresholds or provide a price-elasticity curve.

minor comments (4)

[Section V] The statement that insulated households with HVAC systems 'may lead to cost savings up to 50% approximately' is not directly traceable to any table or figure; please specify the exact comparison and provide the supporting number.
[Section IV] The description 'We run daily simulations with 15-min time steps for one year' combined with 'we use a look-ahead window of one day' is ambiguous: it is unclear how the state is passed between consecutive daily optimizations and whether a full-year receding-horizon MPC is being approximated. Please clarify.
[Section IV and Table II] The comfort light levels (100 and 10000 lux) are stated in the text but not listed in Table II, which otherwise summarizes the comfort constraints; adding them would make the table self-contained.
[Abstract and throughout] The manuscript contains typographical artifacts such as 'increasin g' and 'add ition' in the Abstract; a careful proofread is needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity; the reported savings are simulation outputs of an explicit optimization, not fitted parameters or self-citation-derived conclusions.

full rationale

The paper's central results are obtained by solving the economic MPC problem in Eq. (3) under different comfort bounds defined in Eqs. (5)-(7) and comparing the resulting objective values and load shares. The direction that widening comfort bounds cannot increase the minimum cost is a mathematical property of the optimization (a larger feasible set cannot raise the optimal value), but this is not a circular use of the conclusion: the specific magnitudes, the 16-34% cost savings and the 54.5-57.2% to 64.1-64.7% low-price share shift, are computed outputs rather than assumed inputs. No parameter is fitted to reproduce the headline numbers; the penalty weights are fixed at 1000 and the comfort scenarios are user-defined. The thermal model comes from the external reference [8], and although the auxiliary data reference [18] is by the same authors, it supplies input profiles and matrices, not the paper's conclusions. No uniqueness theorem, ansatz, or fitted quantity is disguised as an independent result. The absence of forecast-error robustness tests is a validity concern, not a circularity concern, because the paper does not claim to demonstrate performance under imperfect forecasts. Thus the derivation chain is self-contained and non-circular.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper's central claims rest on an unvalidated linear building model, perfect forecasts, and hand-chosen penalty weights and thresholds. The model inputs come mostly from prior literature and an auxiliary data link, so the quantitative conclusions inherit those assumptions.

free parameters (3)

Penalty weight for temperature bound violations (rho_temp) = 1000 [EUR/C]
Chosen by hand in Section IV; this weight determines how strongly comfort bounds are enforced as soft constraints, affecting the cost-discomfort trade-off in all results.
Penalty weight for light level violations (rho_l) = 1000 [EUR/lumen]
Chosen by hand in Section IV; not varied in the analysis, so lighting-related results are conditional on this arbitrary value.
Low-price period threshold = 0.5 (normalized price)
Defined in Section IV-A; the price-responsiveness metric (share of power in low-price periods) depends on this threshold, and no sensitivity to it is reported.

assumptions (4)

domain assumption The linear state-space thermal model in Eq. (4) with matrices from [8] and [18] accurately represents the building's heat dynamics after Euler discretization with a 15-minute step.
Introduced in Section III-A; the five states (room, floor, pipe water, refrigerator, water heater) are assumed to follow first-order linear differential equations, neglecting nonlinear effects such as humidity, radiation, and variable COP. All quantitative results depend on this model.
domain assumption Perfect day-ahead knowledge of electricity prices, ambient temperature, solar radiation, occupancy, and hot-water demand.
Stated in Section III-A ('We assume that a set of prices is known a priori') and Section IV-A (one-day look-ahead window). No forecast-error model is included.
domain assumption Occupant comfort is represented by linear penalty functions on temperature and light bounds around a set-point (Eqs. 5-10) with a single penalty weight for all thermal states.
The objective (3) penalizes slack variables linearly; real thermal comfort is nonlinear and asymmetric, so the simulated cost-savings/comfort trade-off is an approximation.
domain assumption The technology parameters (COP, thermal capacities, heat transfer coefficients, appliance powers) are constant and correct for the simulated household.
Values come from prior references [5], [6], [8], [19] and are not validated for this specific building; variations in these parameters would change the numerical results.

how reviews work

0 comments

Cite this review

Pith. "Pith review of How Can Smart Buildings Be Price-Responsive?." pith.science (2026). https://pith.science/paper/F5R5UYJQ

@misc{pith2026190800481,
  author       = {Pith},
  title        = {Pith review of: How Can Smart Buildings Be Price-Responsive?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/F5R5UYJQ}},
  note         = {Machine review of arXiv:1908.00481}
}

read the original abstract

The prospective participation of smart buildings in the electricity system is strongly related to the increasing active role of demand-side resources in the electrical grid. In addition, the growing penetration of smart meters and recent advances on home automation technologies will spur the development of new mathematical tools to help optimize the local resources of these buildings. Within this context, this paper first provides a comprehensive model to determine the electrical consumption of a single-zone household based on economic model predictive control. The goal of this problem is to minimize the electricity consumption cost while accounting for the heating dynamics of the building, smart home appliances, and comfort constraints. This paper then identifies and analyzes the key parameters responsible for the price-responsive behaviour of smart households.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 21 canonical work pages

[18]

Ho w can smart buildings be price-responsive? – Auxiliary data

R. Fern´ andez-Blanco, J. M. Morales, and S. Pineda, “Ho w can smart buildings be price-responsive? – Auxiliary data.” [Online ]. Available: https://goo.gl/WqrhPp, 2019

work page 2019
[8]

Eco- nomic model predictive control for building climate contro l in a smart grid,

R. Halvgaard, N. K. Poulsen, H. Madsen, and J. B. Jorgense n, “Eco- nomic model predictive control for building climate contro l in a smart grid,” 2012 IEEE PES Innovative Smart Grid Technologies (ISGT) , pp. 1–6, 2012

work page 2012
[1]

Demand-side view of electricity market s,

D. S. Kirschen, “Demand-side view of electricity market s,” IEEE Transactions on Power Systems , vol. 18, no. 2, pp. 520–527, May 2003

work page 2003
[2]

Ensuring proﬁtability of energy stor age,

Y . Dvorkin, R. Fern´ andez-Blanco, D. Kirschen, H. Pandzic, J.-P . Watson, and C. Silva-Monroy, “Ensuring proﬁtability of energy stor age,” IEEE Transactions on Power Systems , vol. 32, no. 1, pp. 611–623, Jan. 2017

work page 2017
[3]

Benchmarking smart metering deployment in the EU-27 wi th a focus on electricity COM(2014) 356 ﬁnal,

“Benchmarking smart metering deployment in the EU-27 wi th a focus on electricity COM(2014) 356 ﬁnal,” tech. rep., European Co mmission, Brussels, 2014

work page 2014
[4]

Electricity market design f or the prosumer era,

Y . Parag and B. K. Sovacool, “Electricity market design f or the prosumer era,” Nature Energy, vol. 1, no. 4, p. 16032, 2016

work page 2016
[5]

Grey-box modeling for system identiﬁcation of household r efrigerators: A step toward smart appliances,

G. T. Costanzo, F. Sossan, M. Marinelli, P . Bacher, and H. Madsen, “Grey-box modeling for system identiﬁcation of household r efrigerators: A step toward smart appliances,” IYCE 2013 - 4th International Youth Conference on Energy , 2013

work page 2013
[6]

Scheduling of domestic water heater power demand for maxim izing PV self-consumption using model predictive control,

F. Sossan, A. M. Kosek, S. Martinenas, M. Marinelli, and H . Bindner, “Scheduling of domestic water heater power demand for maxim izing PV self-consumption using model predictive control,” 2013 4th IEEE/PES Innovative Smart Grid Technologies Europe, ISGT Europe 201 3, 2013

work page 2013

Show all 21 references

[7]

Model predictive control for a smart solar tank based on weather and consumption forecasts,

R. Halvgaard et al. , “Model predictive control for a smart solar tank based on weather and consumption forecasts,” Energy Procedia, vol. 30, pp. 270–278, 2012

2012
[9]

Market-based coor dination of thermostatically controlled loads – Part I: A mechanism d esign formulation,

S. Li, W. Zhang, J. Lian, and K. Kalsi, “Market-based coor dination of thermostatically controlled loads – Part I: A mechanism d esign formulation,” IEEE Transactions on Power Systems , vol. 31, no. 2, pp. 1170–1178, 2016

2016
[10]

Braided cobwebs: Cauti onary tales for dynamic pricing in retail electric power markets,

A. G. Thomas and L. Tesfatsion, “Braided cobwebs: Cauti onary tales for dynamic pricing in retail electric power markets,” IEEE Transactions on Power Systems , vol. 33, no. 6, pp. 6870–6882, Nov. 2018

2018
[11]

Short-term forecas ting of price- responsive loads using inverse optimization,

J. Saez-Gallego and J. M. Morales, “Short-term forecas ting of price- responsive loads using inverse optimization,” IEEE Transactions on Smart Grid , vol. 9, no. 5, pp. 4805–4814, Sep. 2018

2018
[12]

A uniﬁed stochastic hybrid system approach to aggregate modeling of responsive loads,

L. Zhao and W. Zhang, “A uniﬁed stochastic hybrid system approach to aggregate modeling of responsive loads,” IEEE Transactions on Automatic Control, vol. 63, pp. 4250–4263, Dec. 2018

2018
[13]

De- centralized coordination of a building manager and an elect ric vehicle aggregator,

J. E. Contreras-Oca˜ na, M. R. Sarker, and M. A. Ortega-V azquez, “De- centralized coordination of a building manager and an elect ric vehicle aggregator,” IEEE Transactions on Smart Grid , vol. 9, no. 4, pp. 2625– 2637, Jul. 2018

2018
[14]

A decentralized fr amework for the optimal coordination of distributed energy resourc es,

M. F. Anjos, A. Lodi, and M. Tanneau, “A decentralized fr amework for the optimal coordination of distributed energy resourc es,” IEEE Transactions on Power Systems , vol. 34, no. 1, pp. 349–359, Jan. 2019

2019
[15]

Smart residential load simulator for energy man- agement in smart grids,

J. M. Gonzalez et al., “Smart residential load simulator for energy man- agement in smart grids,” IEEE Transactions on Industrial Electronics , vol. 66, no. 2, pp. 1443–1452, Feb. 2019

2019
[16]

Characterizing the energy ﬂexibility of buildings and districts,

R. G. Junker et al. , “Characterizing the energy ﬂexibility of buildings and districts,” Applied Energy , vol. 225, pp. 175–182, Sep. 2018

2018
[17]

Halvgaard, J

R. Halvgaard, J. B. Jørgensen, N. Kjølstad, and H. Madse n, Model Predictive Control for Smart Energy Systems . Thesis, Kgs. Lyngby: Technical University of Denmark (DTU), 2014

2014
[19]

Synergy potential of smart a ppliances. Report of the Smart-A project,

R. Stamminger, G. Broil, C. Pakula, H. Jungbecker, M. Br aun, I. R¨ udenauer, and C. Wendker, “Synergy potential of smart a ppliances. Report of the Smart-A project,” tech. rep., 2008

2008
[20]

IBM ILOG CPLEX Optimisation Studio

“IBM ILOG CPLEX Optimisation Studio.” [Online]. Avail able: https://www.ibm.com/analytics/cplex-optimizer, 2019

2019
[21]

Pyomo: Mode ling and solving mathematical programs in Python,

W. E. Hart, J. P . Watson, and D. L. Woodruff, “Pyomo: Mode ling and solving mathematical programs in Python,” Mathematical Programming Computation, vol. 3, no. 3, pp. 219–260, 2011

2011

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [18]

Ho w can smart buildings be price-responsive? – Auxiliary data

R. Fern´ andez-Blanco, J. M. Morales, and S. Pineda, “Ho w can smart buildings be price-responsive? – Auxiliary data.” [Online ]. Available: https://goo.gl/WqrhPp, 2019

work page 2019

[2] [8]

Eco- nomic model predictive control for building climate contro l in a smart grid,

R. Halvgaard, N. K. Poulsen, H. Madsen, and J. B. Jorgense n, “Eco- nomic model predictive control for building climate contro l in a smart grid,” 2012 IEEE PES Innovative Smart Grid Technologies (ISGT) , pp. 1–6, 2012

work page 2012

[3] [1]

Demand-side view of electricity market s,

D. S. Kirschen, “Demand-side view of electricity market s,” IEEE Transactions on Power Systems , vol. 18, no. 2, pp. 520–527, May 2003

work page 2003

[4] [2]

Ensuring proﬁtability of energy stor age,

Y . Dvorkin, R. Fern´ andez-Blanco, D. Kirschen, H. Pandzic, J.-P . Watson, and C. Silva-Monroy, “Ensuring proﬁtability of energy stor age,” IEEE Transactions on Power Systems , vol. 32, no. 1, pp. 611–623, Jan. 2017

work page 2017

[5] [3]

Benchmarking smart metering deployment in the EU-27 wi th a focus on electricity COM(2014) 356 ﬁnal,

“Benchmarking smart metering deployment in the EU-27 wi th a focus on electricity COM(2014) 356 ﬁnal,” tech. rep., European Co mmission, Brussels, 2014

work page 2014

[6] [4]

Electricity market design f or the prosumer era,

Y . Parag and B. K. Sovacool, “Electricity market design f or the prosumer era,” Nature Energy, vol. 1, no. 4, p. 16032, 2016

work page 2016

[7] [5]

Grey-box modeling for system identiﬁcation of household r efrigerators: A step toward smart appliances,

G. T. Costanzo, F. Sossan, M. Marinelli, P . Bacher, and H. Madsen, “Grey-box modeling for system identiﬁcation of household r efrigerators: A step toward smart appliances,” IYCE 2013 - 4th International Youth Conference on Energy , 2013

work page 2013

[8] [6]

Scheduling of domestic water heater power demand for maxim izing PV self-consumption using model predictive control,

F. Sossan, A. M. Kosek, S. Martinenas, M. Marinelli, and H . Bindner, “Scheduling of domestic water heater power demand for maxim izing PV self-consumption using model predictive control,” 2013 4th IEEE/PES Innovative Smart Grid Technologies Europe, ISGT Europe 201 3, 2013

work page 2013