PaperlayerView exact revision on arXiv

Source-linked AI summary

A Review on Energy Consumption Optimization Techniques in IoT Based Smart Building Environments

Abdul Salam Shah, Haidawati Nasir, Muhammad Fayaz, Adidah Lajis, Asadullah Shah

arXiv:1904.09821v1cs.OH

TL;DR

Smart-home energy optimization still faces the challenge of reducing power use while maintaining resident comfort. This paper reviews optimization and scheduling techniques and related comfort and computing topics, concluding that genetic algorithms performed better than other algorithms while several parameters and user preferences remain gaps.

  • Problem

    Smart-home energy optimization must balance resident comfort and power consumption amid environmental variation and unresolved cooling and fuzzy-rule challenges.

    Source: §1 · p. 2

  • Method

    The paper reviews smart-home energy optimization and scheduling, including algorithms, comfort parameters, and edge and fog computing techniques.

    Source: §1 · p. 3

  • Results

    The review concludes that genetic algorithms performed better than other algorithms, while user preferences and several comfort and environmental parameters remain gaps.

    Source: §5 · pp. 19–20

  • Takeaways & Limitations

    The review identifies underconsidered comfort and environmental parameters, including humidity, airflow, heat radiation or flow, cloth thermal insulation, metabolic rate, and water vapor pressure.

    Source: §5 · pp. 19–20

    Fuzzy membership-function parameters are selected manually and cannot be changed by users after initial selection.

    Source: §6 · p. 20

Abstract

from arXiv · show

In recent years, due to the unnecessary wastage of electrical energy in residential buildings, the requirement of energy optimization and user comfort has gained vital importance. In the literature, various techniques have been proposed addressing the energy optimization problem. The goal of each technique was to maintain a balance between user comfort and energy requirements such that the user can achieve the desired comfort level with the minimum amount of energy consumption. Researchers have addressed the issue with the help of different optimization algorithms and variations in the parameters to reduce energy consumption. To the best of our knowledge, this problem is not solved yet due to its challenging nature. The gap in the literature is due to the advancements in the technology and drawbacks of the optimization algorithms and the introduction of different new optimization algorithms. Further, many newly proposed optimization algorithms which have produced better accuracy on the benchmark instances but have not been applied yet for the optimization of energy consumption in smart homes. In this paper, we have carried out a detailed literature review of the techniques used for the optimization of energy consumption and scheduling in smart homes. The detailed discussion has been carried out on different factors contributing towards thermal comfort, visual comfort, and air quality comfort. We have also reviewed the fog and edge computing techniques used in smart homes.

1. Introduction

Smart-home energy optimization must reduce power use while maintaining resident comfort, but environmental variation, whole-building approaches, and fuzzy-rule design leave unresolved challenges. This review surveys optimization and scheduling techniques, with more detailed analysis of optimization.

1. Introduction

  • Traditional forecasting methods use statistical analysis and machine learning; digital-meter data can support hourly prediction as a first step toward optimization.The introduction identifies MAE, MSE, and RMSE as measures used in prior consumption-prediction work.Source: §1 · pp. 1–2
  • Energy optimization aims to reduce consumption while matching appliance power to room conditions, but comfort–power balance, geographic variation, cooling, and fuzzy-rule design remain challenges.The review describes optimization parameters as current indoor conditions and resident-set values; reducing their error difference is linked to lower power consumption.Source: §1 · pp. 2–3
  • Prior studies often optimize whole buildings, while the review identifies individual-area optimization as a gap and notes that upper portions remain warmer than lower portions.The authors state that dividing whole-building optimization into subparts could save more energy.Source: §1 · p. 3
  • The review addresses a stated gap in coverage of smart-home energy optimization and scheduling, identifying techniques, algorithms, comfort parameters, technologies, and edge and fog computing.Although both optimization and scheduling are reviewed, the paper gives optimization the more detailed critical analysis.Source: §1 · p. 3

2. Review Methodology

The methodology frames five research questions, searches four literature databases using defined search strings, and applies eligibility criteria focused on comfort, energy optimization, and related approaches.

2.1. Research Questions

  • The five research questions cover optimization algorithms and parameters, scheduling algorithms and pricing, edge and fog computing, device-connectivity technologies, and comfort-index parameters.The questions are presented as the basis for the mapping study.Source: §2.1 · p. 3

2.2. Searching for Literature

  • The literature search used Web of Science, IEEE Xplore, Google Scholar, and Scopus, with search strings defined for each database.The search criteria were based on factors in the research questions.Source: §2.2 · p. 3, Table 1 · p. 4

2.3. Inclusion/Exclusion of Literature

  • Eligibility focused on comfort-index and energy-optimization studies, including genetic algorithms with PID or fuzzy-logic controllers and work addressing cost reduction or energy savings.The review excluded energy-efficient building-design studies and included peer-reviewed journal and conference articles answering the research questions.Source: §2.3 · p. 4

3. Results

The review screened search results and used forward and backward snowballing to identify 115 qualifying articles. It then conducted detailed analysis of optimization and scheduling subsets.

3. Results

  • The search retrieved 3,765 records from database searches and 35 through Google Scholar before duplicate removal and relevance screening.Full-article review excluded studies that did not consider comfort index or focused on other domains or building structure.Source: §3 · p. 4
  • 115 articles qualified after screening and snowballing, which added 16 papers to the 99 shortlisted articles.The flow diagram summarizes article inclusion and exclusion.Source: §3 · pp. 4–5, Figure 1 · p. 5
  • Detailed analysis covered 36 optimization articles and 20 scheduling articles; the remaining selected studies were formally reviewed to answer the research questions.The paper directs readers to separate appendices for the optimization and scheduling article lists.Source: §3 · p. 5

4. Discussion

The review maps optimization research in smart homes and surveys the algorithms, comfort parameters, and control approaches used to manage energy consumption. Its mapping finds strong emphasis on thermal comfort and single-user settings, while visual and air-quality comfort receive less attention.

4.1. Research Question 1: RQ1: Which algorithms/techniques and parameters have been used for energy consumption optimization in smart homes?

  • The review extracted algorithm, comfort, environment, and user parameters from the articles and used systematic mapping to summarize their frequencies.The optimization map covers 36 articles and reports technique, parameter, and publication-year frequencies.Source: §4.1 · p. 6, §4.1.1 · p. 6

4.1.1. Systematic Mapping of Optimization Techniques

  • 33 of 36 optimization articles considered thermal comfort, 32 considered temperature, and 32 focused on energy optimization, while only 2 considered multi-user comfort.The review also reports that 28 studies considered single-user preferences and 8 optimized individual rooms.Source: §4.1.1 · p. 6, Figure 2 · p. 7
  • Visual comfort and illumination appeared in 19 articles each, while air-quality comfort and CO2 appeared in 16 each.Source: §4.1.1 · p. 6
  • Comfort index was considered in 34 articles and energy saving in 27; genetic algorithms were the most frequently reported optimization algorithm for both objectives.Genetic algorithms appeared in 10 articles for comfort index and 7 for energy saving.Source: §4.1.1 · p. 6

4.1.2. Algorithms / Techniques used for the Energy Optimization

  • Conventional PID controllers face overshoot and tuning issues, motivating fuzzy-control and PID approaches that target comfort and energy use.Source: §4.1.2 · pp. 7–8
  • Reviewed approaches include comfort-zone learning, PMV and LEE thermal indices [36, 37], and multi-agent or predictive control systems.Source: §4.1.2 · pp. 8–9
  • Other studies use motion-sensor lighting, reinforcement learning for occupant-centered lighting, and predictive models coupled with fuzzy control.Source: §4.1.2 · pp. 9–10
  • Reviewed optimization methods also include ant colony optimization [52], particle swarm optimization for individual rooms, and genetic programming with fuzzy logic.Source: §4.1.2 · pp. 9–10

4.2. Research Question 2: RQ2: Which algorithms/techniques, parameters, and pricing schemes have been used for energy consumption scheduling in smart homes?

The review maps scheduling algorithms, parameters, and pricing schemes, and surveys methods for shifting household loads. Across the mapped articles, energy saving and reduced power consumption dominate, while comfort is considered less often.

4.2. Research Question 2: RQ2: Which algorithms/techniques, parameters, and pricing schemes have been used for energy consumption scheduling in smart homes?

  • The scheduling review extracted algorithm, pricing, comfort, and energy-management parameters from 20 articles and mapped their frequencies.Source: §4.2 · p. 10, §4.2.1 · p. 11

4.2.1. Systematic Mapping of Scheduling Techniques

  • Real-time pricing appeared in 9 scheduling articles, critical peak pricing in 8, and day-ahead and time-of-use pricing in 4 each.Flat pricing appeared in 1 article.Source: §4.2.1 · p. 11
  • All 20 scheduling articles focused on energy saving and reducing power consumption; 19 addressed optimal scheduling, 17 peak-to-average ratio reduction, and 15 appliance-based division.Only 4 articles considered comfort index.Source: §4.2.1 · p. 11, Figure 3 · p. 12

4.2.2. Algorithms / Techniques for Energy Optimization through Scheduling

  • Rescheduling appliances from peak to regular hours can lower energy costs, but may disturb comfort and restrict use to low-price hours.Source: §4.2.2 · p. 12
  • Genetic-algorithm demand-side management outperformed heuristic evolutionary algorithms [73], while teacher learning-based optimization outperformed other techniques for optimal scheduling.Source: §4.2.2 · pp. 13–14

4.3. Research Question 3: RQ3: How the edge and fog computing techniques are used in smart homes?

The review surveys edge and fog approaches for smart-home and smart-grid energy management, but notes that these emerging areas have limited literature. It also outlines smart-home technologies and the comfort, automation, and energy-management domains where optimization is considered.

4.3. Research Question 3: RQ3: How the edge and fog computing techniques are used in smart homes?

  • Edge and fog computing are presented as emerging areas with limited literature; the reviewed approaches aim to manage requests and reduce response time.The authors therefore report the most relevant studies available.Source: §4.3 · p. 14, §4.3.1 · p. 16

4.3.1. Edge Computing, Fog Computing based Techniques in Smart Homes/Smart Grid

  • Edge and fog work also addresses smart-home interoperability, distributed home automation, and energy management, with some techniques designed to integrate with edge computing.The reviewed examples include KNN and decision trees for energy management and ZiWi fog-based communication between ZigBee and WiFi devices.Source: §4.3.1 · pp. 14–16
  • Fog-based smart-grid studies place fog layers between cloud systems and grids, using load balancing and virtual-machine allocation to manage requests.Examples include round robin, throttled, beam search, particle swarm optimization, and threshold-based load balancing.Source: §4.3.1 · pp. 14–15

4.4.1. Smart Homes

  • Smart homes connect intelligent devices for monitoring, comfort, energy conservation, maintenance, security, and resident control through displays or mobile devices.The paper describes full-fledged smart homes as not yet a reality and identifies HVAC energy efficiency and user comfort as areas needing improvement.Source: §4.4.1 · p. 16

4.4.2. Technologies Used in Smart Homes

  • Smart-home connectivity uses protocols and networks including Zigbee, Z-Wave, KNX, Ethernet, Wi-Fi, and Bluetooth, while energy-management controllers regulate devices such as HVAC equipment, purifiers, and lights.The paper frames optimization across smart-city IoT domains, with this review focused particularly on smart-home comfort and automation and to a lesser extent energy management.Source: §4.4.2 · p. 17, §4.4.3 · p. 17, Figure 4 · p. 17

4.5. Research Question 5: RQ5: What are the different comfort index parameters in smart homes?

The review treats comfort in smart homes as a range of environmental and human-related parameters, focusing on thermal, visual, and air-quality comfort. It also notes limitations in the review's literature-search scope.

4.5. Research Question 5: RQ5: What are the different comfort index parameters in smart homes?

  • Comfort is framed through thermal, visual, and air-quality parameters within ranges where residents feel comfortable and productivity increases.The study focuses on these three comfort types.Source: §4.5 · p. 18

4.5.1. Thermal Comfort

  • Thermal comfort includes environmental factors such as temperature, humidity, airflow, and heat radiation, alongside human and psychological factors.Human factors include activity, metabolic rate, clothing, gender, age, and season; psychological factors include odor, noise, and visibility.Source: §4.5.1 · p. 18, Figure 5 · p. 18

4.5.2. Visual Comfort

  • Visual comfort combines environmental conditions—including illumination, luminance, glare, contrast, color, and flicker—with individual visual sensitivity, age, and adaptation.Because glare and related conditions are difficult to measure, many researchers use illumination as the visual-comfort measure.Source: §4.5.2 · pp. 18–19, Figure 6 · p. 18

4.5.3. Air Quality Comfort

  • Air-quality comfort is assessed mostly through CO2, with total volatile organic compounds and volatile organic compounds also identified as factors.The passage gives 800 ppm as the desired CO2 concentration and describes purifiers as a way to help maintain the comfort zone.Source: §4.5.3 · p. 19, Figure 7 · p. 19

4.6. Limitations of the study

  • The review's search was limited by its use of relatively few databases and search strings and by excluding literature published before 1996.Source: §4.6 · p. 19

5. Conclusion

The review concludes that genetic algorithms performed better than other algorithms in the studies it considered, while identifying gaps in comfort parameters and user preferences. It also points to newer optimization algorithms as candidates for future energy-optimization research.

5. Conclusion

  • Genetic algorithms performed better than other algorithms in the reviewed residential-building studies addressing comfort and energy reduction.Source: §5 · p. 19
  • The review identifies limited consideration of humidity, multiple users, external conditions, airflow, heat radiation or flow, clothing insulation, metabolic rate, and water vapor pressure.It highlights humidity as important in Malaysia's environmental conditions.Source: §5 · pp. 19–20
  • Earthworm optimization, firefly, bat, anarchic society, cuckoo optimization, league championship, and crow search algorithms are identified as newer candidates for smart-home energy optimization.Some had already been applied to scheduling, and the review notes that relatively few researchers had used crow search for energy optimization.Source: §5 · p. 20

6. Future Work Directions

The future-work discussion identifies gaps in multi-user comfort, adaptive parameter selection, scheduling, and smart-home networking. It also points to fuzzy-controller expertise and fog/edge computing as relevant design considerations.

Adaptive user parameters

  • Automatic membership-function parameter selection could adapt to weather, while current manual settings cannot be changed after initial selection.The authors connect automatic user-parameter prediction with improved comfort and broader accommodation, including for children and people with disabilities.Source: §6 · p. 20

Scheduling and comfort

  • Scheduling can shift appliances to off-peak hours, but the reviewed techniques often compromise user comfort and use waiting time as its measure.The review contrasts this with optimization techniques that consider thermal, visual, and air-quality comfort alongside energy use.Source: §6 · p. 20

Controller design

  • Fuzzy controllers are presented as intuitive to design but dependent on human expertise for successful results.The authors describe them as real-time expert systems that implement human experience and knowledge.Source: §6 · p. 20

Networking

  • Smart-home networking lacks a proper communication protocol, with security concerns and limited compatibility among appliances from different vendors.The authors note that users may need to buy appliances from one vendor to avoid connectivity problems.Source: §6 · p. 20

Fog and edge computing

  • Fog architectures are proposed to reduce delay and response time, while edge computing is described as a developing approach for supplying computing power near the network edges.The discussion places fog between cloud and smart grid and describes edge computing as helping avoid delay.Source: §6 · pp. 20–21

Multi-user comfort

  • Future optimization work should jointly minimize energy use and maximize comfort across residents with different comfort ranges.The authors propose a new or hybrid technique that considers a multi-user comfort index rather than single-user models.Source: §6 · p. 21

Appendix A

Appendix A identifies a table of articles selected for the review and includes fragments of numbered references.

Table A1

  • Table A1 is identified as a listing of articles selected for the review.Source: Table A1 · p. 21

Listed references

  • Reference fragments include Klein et al., Khan et al., and Nassif et al..Source: Appendix A · p. 22
  • Other listed works include Javaid et al., Aslam et al., and Samuel et al..Source: Appendix A · p. 22

Appendix B

Appendix B presents a comparative analysis of energy-optimization techniques, with entries naming methods and comfort or energy-related dimensions.

Table A2

  • Table A2 is titled as a comparative analysis of energy-optimization techniques, and listed dimensions include comfort index and energy saving.Source: Table A2 · p. 22, Appendix B · p. 23

Optimization methods

  • The method labels include Genetic Algorithm (MOGA), Multi-Objective Genetic Algorithm (MOGA), and Hybrid Multi-Objective Genetic Algorithm (HMOGA).Source: Appendix B · pp. 23–24

Controllers

  • The listed approaches also include a Linear Reinforcement Learning Controller, Predictive Control Strategy, and fuzzy controllers.Source: Appendix B · pp. 23–24

Other approaches

  • Other method labels include Artificial Neural Network, Mamdani and Sugeno Fuzzy Inference Systems, and Neural Network with External Inputs.Source: Appendix B · pp. 23–24

Comfort dimensions

  • The comparison's listed outcomes and considerations include thermal comfort, visual comfort, occupant behavior, and multiple occupants' comfort.Source: Appendix B · pp. 23–24

Appendix C

Appendix C identifies a comparative table of energy-scheduling techniques and includes fragments naming methods, goals, and tariff abbreviations.

Table A3

  • Table A3 is identified as a comparative analysis of energy-scheduling techniques.Source: Table A3 · p. 25

Listed methods

  • A listed approach is Action Dependent Heuristic Dynamic Programming (ADHDP), alongside neural networks, backpropagation, and Particle Swarm Optimization.Source: Appendix C · p. 25

Goals and abbreviations

  • The accompanying fragments identify reducing cost and energy saving as goals and list the tariff abbreviations CPP, TOU, FP, APD, and OS.Source: Appendix C · p. 25

RPC RC

The reviewed studies pair energy- and cost-reduction objectives with varied optimization and scheduling methods. The table also lists demand-side management and peak-to-average-ratio reduction among study objectives.

RPC RC

  • Studies [72], [79], [80], and [81] list model predictive control, mixed-integer linear programming, nonlinear optimization, and demand-side scheduling methods.The listed methods include a Hammerstein–Wiener model and temperature setpoint assignment [72], and daily maximum energy scheduling [79, 81].Source: RPC RC · pp. 25–26
  • The cited entries list optimization methods: [73] Genetic Algorithm for energy saving, cost reduction, and demand side management; [74] Earthworm Optimization and Bacterial Foraging; [78] Genetic Algorithm and AFSO; [83] EDE and TLBO; [85] Genetic Algorithm and Earthworm Optimization.Source: RPC RC · pp. 25–26
  • Studies [86], [87], [90], and [92] identify energy-consumption scheduling (ECS), dynamic programming, genetic and binary particle-swarm algorithms, hybrid GAPSO schemes, multiple-knapsack problems (MKP), and a Genetic Harmony Search Algorithm (GHSA).Source: RPC RC · pp. 26–27

Related papers

Loading 1904.09821v1…