Heating, Ventilation, And Air Conditioning Control Research Articles

Automated Control of Transactive HVACs in Energy Distribution Systems

Heating, Ventilation, and Air Conditioning (HVAC) systems contribute significantly to a building's energy consumption. In the recent years, there is an increased interest in developing transactive approaches which could enable automated and flexible scheduling of HVAC systems based on the customer demand and the electricity prices decided by the suppliers. Flexible and automated scheduling of the HVAC systems make it a prime source for participation in residential demand response or transactive energy systems. Therefore, it is of significant interest to identify an optimal strategy to control the HVAC systems. In this article, reducing the energy cost while keeping the comfort level acceptable to the users, we argue that such a control strategy should consider both the energy cost and user comfort simultaneously. Accordingly, we develop the control strategy through the solution of an optimization problem that balances between the energy cost and consumer's dissatisfaction. This optimization enables us to solve a decision-making problem through first price prediction and then choosing HVAC temperature settings throughout the day based on the predicted price, history of the price and HVAC settings, and outside temperature. More specifically, we formulate the control design as a Markov decision process (MDP) using deep neural networks and use Deep Deterministic Policy Gradients (DDPG)-based deep reinforcement learning algorithm to find the optimal control strategy for HVAC systems that balances between electricity cost and user comfort.

What are the impacts on the HVAC system when it provides frequency regulation? – A comprehensive case study with a Multi-Zone variable air volume (VAV) system

Heating, ventilation, and air-conditioning (HVAC) systems in commercial buildings have proven to be a great asset for providing secondary frequency regulation (FR) service to the power grid. However, how the provision of FR service affects the HVAC system performance is still largely unknown. This paper investigates how a HVAC system performance and its associated controls are impacted during and after a FR service. We exploit a dynamic Modelica-based medium office building model as a virtual testbed with the Air Handling Unit (AHU) fans providing the FR service for the assessment. The virtual testbed adopts state-of-the-art HVAC control sequences and an experimentally validated frequency regulation control scheme. The comprehensive study of 832 case studies considers several scenarios, consisting of different regulation capacities (demand levels), standardized FR test signals, and building load profiles in 16 U.S, climate zones. The holistic assessment includes the interaction between the FR and the HVAC baseline controls, the impact of FR on the occupant’s thermal comfort, and, more importantly, the energy efficiency.The results show that FR has little or no impact on the HVAC operation and occupant thermal comfort while the building is providing the service if an appropriate regulation capacity is pre-determined. However, if the regulation capacity is overestimated, the quality of the FR service provided by the HVAC system and the HVAC energy efficiency will be affected due to the equipment operating constraints. In addition, the AHU supply air temperature fluctuates, and the controller cannot maintain the desired setpoint as the FR magnitude is increased. After the FR service time, the system takes a certain period to return to normal operation. The rebound period extends with the increase in the FR magnitude. For this case study, there exists a control conflict between the FR control and the existing HVAC control (from ASHRAE Guideline 36-2018) during high load conditions, causing the fan speed to saturate. The increased fan electricity consumption caused by this control conflict will offset the potential revenue a building can earn from offering the FR service. In addition, the rebound power usage could have a negative effect on the grid.

An Integrated Approach to Adaptive Control and Supervisory Optimisation of HVAC Control Systems for Demand Response Applications

Heating, ventilating, and air-conditioning (HVAC) systems account for a large percentage of energy consumption in buildings. Implementation of efficient optimisation and control mechanisms has been identified as one crucial way to help reduce and shift HVAC systems’ energy consumption to both save economic costs and foster improved integration with renewables. This has led to the development of various control techniques, some of which have produced promising results. However, very few of these control mechanisms have fully considered important factors such as electricity time of use (TOU) price information, occupant thermal comfort, computational complexity, and nonlinear HVAC dynamics to design a demand response schema. In this paper, a novel two-stage integrated approach for such is proposed and evaluated. A model predictive control (MPC)-based optimiser for supervisory setpoint control is integrated with a digital parameter-adaptive controller for use in a demand response/demand management environment. The optimiser is designed to shift the heating load (and hence electrical load) to off-peak periods by minimising a trade-off between thermal comfort and electricity costs, generating a setpoint trajectory for the inner loop HVAC tracking controller. The tracking controller provides HVAC model information to the outer loop for calibration purposes. By way of calibrated simulations, it was found that significant energy saving and cost reduction could be achieved in comparison to a traditional on/off or variable HVAC control system with a fixed setpoint temperature.

Multi-Objective Optimisation-Based Design of an Electric Vehicle Cabin Heating Control System for Improved Thermal Comfort and Driving Range

Modern electric vehicle heating, ventilation, and air-conditioning (HVAC) systems operate in more efficient heat pump mode, thus, improving the driving range under cold ambient conditions. Coupling those HVAC systems with novel heating technologies such as infrared heating panels (IRP) results in a complex system with multiple actuators, which needs to be optimally coordinated to maximise the efficiency and comfort. The paper presents a multi-objective genetic algorithm-based control input allocation method, which relies on a multi-physical HVAC model and a CFD-evaluated cabin airflow distribution model implemented in Dymola. The considered control inputs include the cabin inlet air temperature, blower and radiator fan air mass flows, secondary coolant loop pump speeds, and IRP control settings. The optimisation objective is to minimise total electric power consumption and thermal comfort described by predictive mean vote (PMV) index. Optimisation results indicate that HVAC and IRP controls are effectively decoupled, and that a significant reduction of power consumption (typically from 20% to 30%) can be achieved using IRPs while maintaining the same level of thermal comfort. The previously proposed hierarchical HVAC control strategy is parameterised and extended with a PMV-based controller acting via IRP control inputs. The performance is verified through simulations in a heat-up scenario, and the power consumption reduction potential is analysed for different cabin air temperature setpoints.

Building HVAC control with reinforcement learning for reduction of energy cost and demand charge

Energy efficiency remains a significant topic in the control of building heating, ventilation, and air-conditioning (HVAC) systems, and diverse set of control strategies have been developed to optimize performance, including recently emerging techniques of deep reinforcement learning (DRL). While most existing works have focused on minimizing energy consumption, the generalization to energy cost minimization under time-varying electricity price profiles and demand charges has rarely been studied. Under these utility structures, significant cost savings can be achieved by pre-cooling buildings in the early morning when electricity is cheaper, thereby reducing expensive afternoon consumption and lowering peak demand. However, correctly identifying these savings requires planning horizons of one day or more. To tackle this problem, we develop Deep Q-Network (DQN) with an action processor, defining the environment as a Partially Observable Markov Decision Process (POMDP) with a reward function consisting of energy cost (time-of-use and peak demand charges) and a discomfort penalty, which is an extension of most reward functions used in existing DRL works in this area. Moreover, we develop a reward shaping technique to overcome the issue of reward sparsity caused by the demand charge. Through a single-zone building simulation platform, we demonstrate that the customized DQN outperforms the baseline rule-based policy, saving close to 6% of total cost with demand charges, while close to 8% without demand charges.

Hybrid-Model-Based Deep Reinforcement Learning for Heating, Ventilation, and Air-Conditioning Control

Buildings account for a large proportion of the total energy consumption in many countries and almost half of the energy consumption is caused by the Heating, Ventilation, and air-conditioning (HVAC) systems. The model predictive control of HVAC is a complex task due to the dynamic property of the system and environment, such as temperature and electricity price. Deep reinforcement learning (DRL) is a model-free method that utilizes the “trial and error” mechanism to learn the optimal policy. However, the learning efficiency and learning cost are the main obstacles of the DRL method to practice. To overcome this problem, the hybrid-model-based DRL method is proposed for the HVAC control problem. Firstly, a specific MDPs is defined by considering the energy cost, temperature violation, and action violation. Then the hybrid-model-based DRL method is proposed, which utilizes both the knowledge-driven model and the data-driven model during the whole learning process. Finally, the protection mechanism and adjusting reward methods are used to further reduce the learning cost. The proposed method is tested in a simulation environment using the Australian Energy Market Operator (AEMO) electricity price data and New South Wales temperature data. Simulation results show that 1) the DRL method can reduce the energy cost while maintaining the temperature satisfactory compared to the short term MPC method; 2) the proposed method improves the learning efficiency and reduces the learning cost during the learning process compared to the model-free method.

Distributed Control of Multizone HVAC Systems Considering Indoor Air Quality

This article studies a scalable control method for multizone heating, ventilation, and air-conditioning (HVAC) systems to optimize the energy cost for maintaining thermal comfort (TC) and indoor air quality (IAQ) (represented by CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ) simultaneously. This problem is computationally challenging due to the complex system dynamics, various spatial and temporal couplings, as well as multiple control variables to be coordinated. To address the challenges, we propose a two-level distributed method (TLDM) with an upper level and lower level control integrated. The upper level computes zone mass flow rates for maintaining zone TC with minimal energy cost, and then, the lower level strategically regulates zone mass flow rates and the ventilation rate to achieve IAQ while preserving the near energy-saving performance of upper level. As both the upper and the lower level computation are deployed in a distributed manner, the proposed method is scalable and computationally efficient. The near-optimal performance of the method in energy cost saving is demonstrated through comparison with the centralized method. In addition, the comparisons with the existing distributed method show that our method can provide IAQ with only little increase of energy cost, while the latter fails. Moreover, we demonstrate that our method outperforms the demand-controlled ventilation (DCVs) strategies for IAQ management with about 8%–10% energy cost reduction. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners:</i> The high portion of building energy consumption has motivated the energy saving for heating, ventilation, and air-conditioning (HVAC) systems. Concurrently, the living standards for indoor environment are rising among the occupants. Nevertheless, the status quo on improving building energy efficiency has mostly focused on maintaining thermal comfort (such as temperature), and the indoor air quality (IAQ) (usually represented by CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> level) has been seldom incorporated. In our previous work with the similar setting, we observed that the CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> levels will surge beyond tolerance during the high occupancy periods if only thermal comfort (TC) is considered for HVAC control. This deduces the IAQ and TC should be jointly considered while pursuing the energy cost saving target and thus studied in this article. This task is computationally cumbersome due to the complex system dynamics (thermal and CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ) and tight correlations among the different control components (variable air volume and fresh air damper). To cope with these challenges, this work develops a two-level distributed computation paradigm for HVAC systems based on problem structures. Specifically, the upper level control (ULC) first calculates zone mass flow rates for maintaining comfortable zone temperature with minimal energy cost, and then, the lower level strategically regulates the computed zone mass flow rates as well as ventilation rate to satisfy IAQ while preserving the near energy-saving performance of the ULC. As both the upper and lower level calculations can be implemented in a distributed manner, the proposed method is scalable to large multizone deployment. The method’s performance both in maintaining comfort (i.e., TC and IAQ) and energy cost saving is demonstrated via simulations in comparisons with the centralized method, the distributed token-based scheduling strategy, and the demand-controlled ventilation strategies.

Model-Based and Data-Driven HVAC Control Strategies for Residential Demand Response

The implementations of residential demand response (DR) based on heating, ventilation, and air conditioning (HVAC) are inseparable from effective control algorithms for coordinating the operating schedules of multiple HVAC devices. In this work, both model-based and data-driven HVAC control strategies are developed to determine the optimal control actions for HVAC systems. The control objectives are to minimize customers' electricity costs, customers' discomfort, and the utility-level load violation. In the model-based approach, a thermal resistance-capacitance (RC) HVAC model is formulated to capture buildings' thermodynamic behaviors, and a distributed solution algorithm (i.e., alternating direction method of multipliers) is applied to determine the day-ahead HVAC operation schedules. In the data-driven approach, the neural networks continuously interact with the environment during the training process to learn what control actions to take under certain circumstances and then are used for online decision-making. The case study is performed on a utility system with one hundred houses. Simulation results demonstrate that the model-based approach can save 22% of the total cost compared to the data-driven approach, while the data-driven approach does not require outdoor temperature forecast information and its computational speed is 46 times faster than that of the model-based approach.

Gnu-RL: A Practical and Scalable Reinforcement Learning Solution for Building HVAC Control Using a Differentiable MPC Policy

We propose Gnu-RL: a novel approach that enables real-world deployment of reinforcement learning (RL) for building control and requires no prior information other than historical data from existing Heating Ventilation and Air Conditioning (HVAC) controllers. In contrast with existing RL agents, which are opaque to expert interrogation and need ample training to achieve reasonable control performance, Gnu-RL is much more attractive for real-world applications. Furthermore, Gnu-RL avoids the need to develop and calibrate simulation environments for each building or system under control, thus making it highly scalable. To achieve this, we bootstrap the Gnu-RL agent with domain knowledge and expert demonstration. Specifically, Gnu-RL adopts a recently-developed Differentiable Model Predictive Control (MPC) policy, which encodes domain knowledge on planning and system dynamics, making it both sample-efficient and interpretable. Prior to any interaction with the environment, a Gnu-RL agent is pre-trained on historical data using imitation learning, enabling it to match the behavior of the existing controller. Once it is put in charge of controlling the environment, the agent continues to improve its policy end-to-end, using a policy gradient algorithm. We evaluate Gnu-RL in both simulation studies and a real-world testbed. Firstly, we validated the Gnu-RL agent is indeed practical and scalable by applying it to three commercial reference buildings in simulation, and demonstrated the superiority of the Differentiable MPC policy compared to a generic neural network. In another simulation experiment, our approach saved 6.6% of the total energy compared to the best published RL result for the same environment, while maintaining a higher level of occupant comfort. Finally, Gnu-RL was deployed to control the airflow to a real-world conference room with a Variable-Air-Volume (VAV) system during a three-week period. Our results show that Gnu-RL reduced cooling demand by 16.7% compared to the existing controller and tracked the temperature set-point better.

Optimal HVAC Control for Demand Response via Chance-Constrained Two-Stage Stochastic Program

This study is interested in the application of demand response (DR) to building energy management via heating, ventilation and air conditioning (HVAC) controls. The main objective of this study is to develop an optimization model to appropriately control HVAC operations to minimize cost in response to time-varying electricity prices and intermittent solar power generation. The main idea is to adjust electricity consumption associated with HVAC operations while ensuring the thermal comfort of occupants. In particular, we propose a two-stage stochastic program formulated to find an optimal HVAC operations schedule that comprises proactive and reactive controls to minimize energy cost against the stochastic nature in electricity prices, solar power generation, and weather conditions. Specifically, we introduce a chance constraint that can be appropriately integrated with the proposed optimization model to ensure thermal comfort at the desired level. To validate and evaluate the proposed approach, numerical experiments are designed to compare the proposed approach with the benchmark control policies designed to simulate the practice of the general purpose thermostat and the algorithm proposed by relevant study with various parameter settings. The numerical experiments show that the proposed approach results in significant cost savings while capturing optimal HVAC operations.

Intelligent multi-zone residential HVAC control strategy based on deep reinforcement learning

Residential heating, ventilation, and air conditioning (HVAC) has been considered as an important demand response resource. However, the optimization of residential HVAC control is no trivial task due to the complexity of the thermal dynamic models of buildings and uncertainty associated with both occupant-driven heat loads and weather forecasts. In this paper, we apply a novel model-free deep reinforcement learning (RL) method, known as the deep deterministic policy gradient (DDPG), to generate an optimal control strategy for a multi-zone residential HVAC system with the goal of minimizing energy consumption cost while maintaining the users’ comfort. The applied deep RL-based method learns through continuous interaction with a simulated building environment and without referring to any prior model knowledge. Simulation results show that compared with the state-of-art deep Q network (DQN), the DDPG-based HVAC control strategy can reduce the energy consumption cost by 15% and reduce the comfort violation by 79%; and when compared with a rule-based HVAC control strategy, the comfort violation can be reduced by 98%. In addition, experiments with different building models and retail price models demonstrate that the well-trained DDPG-based HVAC control strategy has high generalization and adaptability to unseen environments, which indicates its practicability for real-world implementation.

Development and Evaluation of Occupancy-Aware HVAC Control for Residential Building Energy Efficiency and Occupant Comfort

Occupancy-aware heating, ventilation, and air conditioning (HVAC) control offers the opportunity to reduce energy use without sacrificing thermal comfort. Residential HVAC systems often use manually-adjusted or constant setpoint temperatures, which heat and cool the house regardless of whether it is needed. By incorporating occupancy-awareness into HVAC control, heating and cooling can be used for only those time periods it is needed. Yet, bringing this technology to fruition is dependent on accurately predicting occupancy. Non-probabilistic prediction models offer an opportunity to use collected occupancy data to predict future occupancy profiles. Smart devices, such as a connected thermostat, which already include occupancy sensors, can be used to provide a continually growing collection of data that can then be harnessed for short-term occupancy prediction by compiling and creating a binary occupancy prediction. Real occupancy data from six homes located in Colorado is analyzed and investigated using this occupancy prediction model. Results show that non-probabilistic occupancy models in combination with occupancy sensors can be combined to provide a hybrid HVAC control with savings on average of 5.0% and without degradation of thermal comfort. Model predictive control provides further opportunities, with the ability to adjust the relative importance between thermal comfort and energy savings to achieve savings between 1% and 13.3% depending on the relative weighting between thermal comfort and energy savings. In all cases, occupancy prediction allows the opportunity for a more intelligent and optimized strategy to residential HVAC control.

Influence of solar radiation on the energy consumption of large buildings on a university campus

This paper reports the numerical study on the influence of solar radiation on the energy consumption of large buildings on a university campus. The actual campus is located in the south of Portugal, in a Mediterranean type environment, and consists of 6 educational buildings. These six buildings have a total area of 27,599 m2 and 595 compartments, where 6,529 opaque surfaces (doors, walls, etc.) and 983 transparent ones (windows) were identified. This study aims to assess numerically how solar radiation transmitted on windows affects the energy consumption of the Heating, Ventilation, and Air-Conditioning (HVAC) systems, controlled by the PMV (Predicted Mean Vote) index, of each of these buildings, and the thermal comfort level of the occupants. Software developed by the authors is used to simulate the thermal behavior of buildings with complex topology. This software evaluates indoor air quality inside the spaces, thermal comfort of the occupants, thermal energy consumption of the HVAC system, and solar radiation distribution outside the buildings and inside the compartments, among others. The HVAC control system based on the PMV index applied in this work was designed to maintain the PMV comfort index within category C of ISO 7730, with a maximum of 15% of people dissatisfied. In order to evaluate the indoor comfort level of the occupants, the totals of cold and warm uncomfortable hours were calculated. Two different weather conditions, typical of the region, were set as inputs for the simulation performed in this study: a typical winter day, and a typical summer day. The outputs obtained were the daily evolution of total solar radiation transmitted on windows, total uncomfortable hours for the occupants, and total HVAC system energy consumption for each building. The results obtained show that, for typical winter conditions, an increase in the transmitted solar radiation on windows causes a decrease in HVAC system energy consumption, and also in the number of uncomfortable hours, which is a favorable situation. On the other hand, for typical summer conditions, it is observed that when transmitted solar radiation on windows increases, HVAC system energy consumption, and the total number of uncomfortable hours increase as well, configuring an unfavorable situation. It is also found that the values of solar radiation transmitted on windows are higher in winter than in summer conditions. In summer, the lowest values of solar radiation transmitted on windows occur at noon. The last two observations lead to the conclusion that, overall, these buildings have correctly positioned passive shading elements, a technique that contributes to an adequate solar passive architectural design.

Building energy management decision-making in the real world: A comparative study of HVAC cooling strategies

In recent years, various heating, ventilation, and air-conditioning (HVAC) control strategies have been proposed to reduce buildings’ energy consumption and environmental impacts. Due to different research designs, it is impossible to directly compare the reported performance of these strategies and determine the best approach for operating real-world buildings. To fill this gap, this research uses a well-validated building simulation model to holistically evaluate the performance of 12 existing HVAC cooling strategies in reducing the total cooling energy use, peak demand, and cooling energy cost. The findings include: i) The ON/OFF and night setup control strategies could lead to even higher energy cost than the 24/7 operation baseline; ii) demand-limiting control methods performed well across the three energy performance metrics and achieved the highest energy reduction ratios ranging from 9.8% to 10.5%; iii) precooling and extended precooling reduced the peak load most significantly by 15.4% and 21.4%, respectively; and iv) the resistive-capacitive (RC) network-based precooling optimization resulted in the highest electricity cost savings with reduction ratios of 16.6% and 12.9%, based on the two common price schedules. Besides providing valuable insights to the research community, this study offers a practical method to help building operators analyze and select the best HVAC control strategies for their energy management goals.

Coupling simulation with artificial neural networks for the optimisation of HVAC controls in manufacturing environments

Manufacturing remains one of the most energy intensive sectors, additionally, the energy used within buildings for heating, ventilation and air conditioning (HVAC) is responsible for almost half of the UK’s energy demand. Commonly, these are analysed in isolation from one another. Use of machine learning is gaining popularity due to its ability to solve non-linear problems with large data sets and little knowledge about relationships between parameters. Such models use relationships between inputs and outputs to make further predictions on unseen data, without requiring any understanding regarding the system, making them highly suited to dealing with the stochastic data sets found in a manufacturing environment. This has been seen in literature for determining electrical energy demand for residential or commercial buildings, rather than manufacturing environments. This study proposes a novel method of coupling simulation with machine learning to predict indoor workshop conditions and building energy demand, in response to production schedules, outdoor conditions, building behaviour and use. Such predictions can subsequently allow for more efficient management of HVAC systems. Based upon predicted energy consumption, potential spikes were identified and manufacturing schedules subsequently optimised to reduce peak energy demand. Coupling simulation techniques with machine learning algorithms eliminates the requirement for costly and intrusive methods of data collection, providing a method of predicting and optimising building energy consumption in the manufacturing sector.

Optimisation of Control Input Allocation Maps for Electric Vehicle Heat Pump-based Cabin Heating Systems

The heating, ventilation and air conditioning (HVAC) system negatively affects the electric vehicle (EV) driving range, especially under cold ambient conditions. Modern HVAC systems based on the vapour-compression cycle can be rearranged to operate in the heat pump mode to improve the overall system efficiency compared to conventional electrical/resistive heaters. Since such an HVAC system is typically equipped with multiple actuators (compressor, pumps, fans, valves), with the majority of them being controlled in open loop, an optimisation-based control input allocation is necessary to achieve the highest efficiency. This paper presents a genetic algorithm optimisation-based HVAC control input allocation method, which utilises a multi-physical HVAC system model implemented in Dymola/Modelica. The considered control inputs include the cabin inlet air temperature reference, blower and radiator fan air mass flows and secondary coolant loop pumps’ speeds. The optimal allocation is subject to specified, target cabin air temperatures and heating power. Additional constraints include actuator hardware limits and safety functions, such as maintaining the superheat temperature at its reference level. The optimisation objective is to maximise the system efficiency defined by the coefficient of performance (COP). The optimised allocation maps are fitted by proper mathematical functions to facilitate the control strategy implementation and calibration. The overall control strategy consists of superimposed cabin air temperature controller that commands heating power, control input allocation functions, and low-level controllers that ensure cabin inlet air and superheat temperature regulation. The control system performance is verified through Dymola simulations for the heat pump mode in a heat-up scenario. Control input allocation map optimisation results are presented for air-conditioning (A/C) mode, as well.

Evaluating the Adaptability of Reinforcement Learning Based HVAC Control for Residential Houses

Intelligent Heating, Ventilation, and Air Conditioning (HVAC) control using deep reinforcement learning (DRL) has recently gained a lot of attention due to its ability to optimally control the complex behavior of the HVAC system. However, more exploration is needed on understanding the adaptability challenges that the DRL agent could face during the deployment phase. Using online learning for such applications is not realistic due to the long learning period and likely poor comfort control during the learning process. Alternatively, DRL can be pre-trained using a building model prior to deployment. However, developing an accurate building model for every house and deploying a pre-trained DRL model for HVAC control would not be cost-effective. In this study, we focus on evaluating the ability of DRL-based HVAC control to provide cost savings when pre-trained on one building model and deployed on different house models with varying user comforts. We observed around 30% of cost reduction by pre-trained model over baseline when validated in a simulation environment and achieved up to 21% cost reduction when deployed in the real house. This finding provides experimental evidence that the pre-trained DRL has the potential to adapt to different house environments and comfort settings.

Nationwide HVAC energy-saving potential quantification for office buildings with occupant-centric controls in various climates

The occupant-centric control (OCC) is receiving an increasing attention since it could reduce building heating ventilation and air-conditioning (HVAC) system energy consumptions while not affecting the occupant thermal comfort. This paper aims to quantify the nationwide energy-saving potential of implementing the occupant-centric HVAC controls in typical office buildings. First, the medium office and large office from the Department of Energy (DOE) Commercial Prototype Building Models (CPBM) were enhanced to have detailed layouts and dynamic occupancy schedules. Then, a comprehensive simulation plan was created by incorporating the multiple zone-level and system-level occupant-centric building HVAC controls recommended by the updated ASHRAE Standard 90.1 – 2019 and ASHRAE Guideline 36 – 2018. Three control scenarios with different occupancy sensing methods were identified in this simulation plan. A nation-wide parametric analysis, which includes two building types, three occupancy sensing scenarios, two building code versions, and 16 U.S. climate zones, was carried out. The simulation results of the key control variables and HVAC energy consumption suggest that generally, both the occupancy presence sensor and occupant counting sensor could achieve energy savings for the office buildings in the majority of the scenarios. However, compared with the occupancy presence sensor, which could support both the temperature setpoint reset and operational breathing zone airflow rate reset for the unoccupied zones, the occupant counting sensor only brings a marginal benefit. Besides, a higher HVAC energy-saving ratio could be achieved in the heating-dominated zone, since the energy reduction brought with the minimum outdoor airflow rate reset is stronger in the heating mode.

Reinforcement learning for whole-building HVAC control and demand response

This paper proposes a novel reinforcement learning (RL) architecture for the efficient scheduling and control of the heating, ventilation and air conditioning (HVAC) system in a commercial building while harnessing its demand response (DR) potentials. With advances in automated building management systems, this can be achieved seamlessly by a smart autonomous RL agent which takes the best action, for example, a change in HVAC temperature set point, necessary to change the electricity usage pattern of a building in response to demand response signals, and with minimal thermal comfort impact to customers. Previous research in this area has tackled only individual aspects of the problem using RL. Specifically, due to the challenges in implementing demand response with whole-building models, simpler analytical models which poorly capture reality have been used instead. And where whole-building models are applied, RL is used for HVAC control mainly to achieve energy efficiency goals while demand response is neglected. Thus, in this research, we implement a holistic framework by designing an efficient RL controller for a whole-building model which learns to optimise and control the HVAC system for improved energy efficiency and thermal comfort levels in addition to achieving demand response goals. Our simulation results show that by applying reinforcement learning for normal HVAC operation, a maximum weekly energy reduction of up to 22% can be achieved compared to a handcrafted baseline controller. Furthermore, by employing a DR-aware RL controller during demand response periods, average power reductions or increases of up to 50% can be achieved on a weekly basis compared to the default RL controller, while keeping occupant thermal comfort levels within acceptable bounds.

Development of CNN-based visual recognition air conditioner for smart buildings

Demand-driven heating, ventilation, and air conditioning (HVAC) operations have become very attractive in energy-efficient smart buildings. Demand-oriented HVAC control largely relies on accurate detection of building occupancy levels and locations. So far, existing building occupancy detection methods have their disadvantages, and cannot fully meet the expected performance. To address this challenge, this paper proposes a visual recognition method based on convolutional neural networks (CNN), which can intelligently interpret visual contents of surveillance cameras to identify the number of occupants and their locations in buildings. The proposed study can detect the quantity, distance, and angle of indoor human users, which is essential for controlling air-conditioners to adjust the direction and speed of air blow. Compared with the state of the art, the proposed method successfully fulfills the function of building occupant counting, which cannot be realized when using PIR, sound, and carbon dioxide sensors. Our method also achieves higher accuracy in detecting moving or stationary human bodies and can filter out false detections (such as animal pets or moving curtains) that are existed in previous solutions. The proposed idea has been implemented and collaboratively tested with air conditioners in an office environment. The experimental results verify the validity and benefits of our proposed idea.