HVAC system optimization with CO2 concentration control using genetic algorithms

doi:10.1016/j.enbuild.2008.12.004

Energy and Buildings

Volume 41, Issue 5, May 2009, Pages 571-577

https://doi.org/10.1016/j.enbuild.2008.12.004 Get rights and content

Abstract

This study describes the use of genetic algorithms (GAs) for operating standard HVAC systems (HVAC—heating, ventilation and air conditioning) in order to optimize performance, primarily with regard to power saving. Genetic algorithms were introduced as an instrument for solving optimization problems. Analytic optimization procedures are widely used in other fields of engineering, but they are difficult to operate within HVAC systems, because the range of the research is usually too broad, the problems are not linear but rather discontinuous, and they mostly have complex limitations. This is why for this type of system genetic algorithms are used, since they have the qualities of robustness and efficiency that are crucial for finding the optimal solution. A simulation is conducted in order to demonstrate how much power can be saved by using the suggested method of CO₂ concentration control in a standard HVAC system. In addition to Matlab Simulink, the suggested method is verified with Energy software.

Introduction

For the past few years, HVAC systems have been extremely popular in business office buildings where their prime role is to provide convenient working conditions. Convenient working conditions include desirable air temperature, the required amount of oxygen and carbon dioxide, regular circulation of the air in the offices, etc.

For satisfying such requirements, it is necessary to implement an efficient and high-quality operating mode for the HVAC system that would not only answer all the demands mentioned, but would take energy efficiency into account as well. This means it is necessary to fulfill all these tasks using as little energy as possible. There are a great number of different methodologies that describe the design of controllers for the exact purpose of operating HVAC systems. One of the possibilities is the use of genetic algorithms, due to their very favorable characteristics and the wide range of problems they cover [23].

Genetic algorithms are very good at finding the optimal solution among the appropriate scope of solutions they search [24]. Genetic algorithms are search algorithms based upon the principles of Darwin's theory of natural selection. The basic operations that genetic algorithms utilize are reproduction, hybridization, mutation and selection. Reproduction ensures the continuation of the best individuals, that is to say, the individuals with the highest fitness, from the present to the next generation. Hybridization combines the genes of two parental individuals and produces a completely unique individual. Mutation changes the structure of genes of each individual separately. Using the previous operations, new individuals are being derived upon whom a selection will be conducted with the aim of choosing the best individuals, i.e. the individuals with the best features that will directly be passed into the next generation. The quality of each individual is presented through its fitness function used by the operation of selection.

The theory of genetic algorithms and their application to HVAC optimization is presented by Lu et al. [1], Atthajariyakul and Leephakpreeda [2], Huang and Lam [3], Asiedu et al. [4], Chow et al. [5].

Section snippets

Genetic algorithms

Genetic algorithms are powerful general purpose stochastic optimization methods which have been inspired by the Darwinian evolution of a population subject to reproduction, crossover the mutations in a selective environment where the fitness survive. GA combines the artificial survival of the fitness with genetic operators abstracted from nature to form a very robust mechanism that is suitable for a variety of optimization problems. In mathematical terms the goal of genetic algorithm is to

Description of HVAC systems and principle of work

In Fig. 1, the HVAC system discussed in this paper is symbolically presented. It consists of several functional components described.

The system consists of supply tubes through which the outside and the return air are delivered, a mixing box for mixing of these two airs, a cooling coil, a chiller, a three-way valve, a blower, outside and return air dampers, and a one-speed pump which supplies a permanent flow of water through the chiller. The valve and the dampers contain pneumatic propelling

Model of HVAC systems

On the modeling of HVAC systems, steady-state models have been largely presented such as in Knabe and Le [6], Hensen [7], Chow et al. [8], Bourdouxhe and André [9], Lam et al. [10], Cui et al. [11]. On the other hand, unsteady-state mathematical models have been presented by Novak et al. [12] and Barbosa and Mendes [13].

The mathematical model of the HVAC system used in this paper, with its functional scheme given in Fig. 1, is completely realized in Matlab's tool ‘Simulink’ (Anderson et al. [14]

Simulation results using Matlab Simulink

The simulation was conducted in such manner that all the input parameters into the simulink model were redefined and returned into that model, while only for the input of C_dr such values are used that would entirely open the outer air damper (case 1), or values were used that were derived through a previous calculation by means of a genetic algorithm (cases 2, 3 and 4). For that reason, for the purpose of the simulation, the following values are set at the initiation of the model: (Table 1).

The

Result verification using EnergyPlus software

EnergyPlus, which is widely accepted as a tool for simulation, enables checking new solutions on real building models in a very simple way. In a similar way, a multi-zone airflow model was confirmed by Huang et al. [15], building heat balance by Strand et al. [16] and other simulations by Fisher et al. [17], Brent and Ellis [18], Strand and Baumgartner [19], Zhou et al. [20].

Based on previously mentioned models, a detailed model of a business building in Belgrade (Fig. 5) was analyzed using the

Conclusion

If we accept, for example, that 1 m³ of cold water cooled by our chiller costs 0.5 euro (€), then recalculating the total daily flow (from 08.00 a.m. to 05.00 p.m. during 1 day) in all four cases, we get a daily sum (calculated in euros) that is necessary to cover the expense for cooling the water with the chiller. In the case where the genetic algorithm is not applied, 14.25 m³ of cold water flows through the cooling coil. In that case the price of cooling is 7.12€ on a daily basis. The flow of

References (24)

S. Atthajariyakul et al.
Neural computing thermal comfort index for HVAC systems
Energy Conversion and Management
(2005)
W. Huang et al.
Using genetic algorithms to optimize controller parameters for HVAC systems
Energy and Buildings
(1997)
T.T. Chow et al.
Global optimization of absorption chiller system by genetic algorithm and neural network
Energy and Buildings
(2002)
T.T. Chow et al.
Primitive parts: an approach to air-conditioning component modeling
Energy and Buildings
(1997)
M. Anderson et al.
An experimental system for advanced heating, ventilating and air conditioning (HVAC) control
Energy and Buildings
(2007)
R.K. Strand et al.
Modeling radiant heating and cooling systems: integration with a whole-building simulation program
Energy and Buildings
(2005)
Y.P. Zhou et al.
Energy simulation in the variable refrigerant flow air-conditioning system under cooling conditions
Energy and Buildings
(2007)
L. Lu et al.
HVAC system optimisation—in-building section
Energy and Buildings
(2005)
Y. Asiedu et al.
HVAC duct system design using genetic algorithms
HVAC&R Research
(2000)
G. Knabe et al.
Building simulation by application of a HVAC system considering the thermal and moisture behaviors of the perimeter walls

J.L.M. Hensen, On the thermal interaction of building structure and heating and ventilating system, PhD Thesis,...

J.P. Bourdouxhe et al.

Simulation of a centralized cooling plant under different control strategies

Cited by (112)

MF^2: Model-free reinforcement learning for modeling-free building HVAC control with data-driven environment construction in a residential building
2023, Building and Environment
Reinforcement Learning (RL) has advanced energy-efficient control of building Heating, Ventilation and Air Conditioning (HVAC) systems. Constructing a suitable RL environment for buildings is a crucial challenge. Compared to widely-used simulation-based environments, data-driven approaches offer higher training efficiency but face convergence difficulties due to influential factors, limiting their current application.
To explore data-driven construction of RL environments for building HVAC systems, this study proposes two strategies for controlling room temperature setpoints in a residential building. XGBoost and Long Short-Term Memory Network (LSTM) are trained for energy consumption and room temperature prediction. One strategy predicts parameters for on-off states, while the other for power-on states. The XGBoost models are integrated into an OpenAI Gym environment. The first strategy achieves 0.8634 R² and 0.2423 Root Mean Squared Error (RMSE) for energy consumption prediction. The R² of room air temperature models are approximately 0.99 and the RMSE are lower than 0.31. The second strategy achieves 0.9181 R² and 0.1042 RMSE for energy consumption prediction and similar performance for room temperature prediction. Deep Q-learning (DQN) and Deep Deterministic Policy Gradient (DDPG) algorithms are separately trained using these environments. Results show that the first strategy fails to induce the correct training of RL models, while the second strategy successfully induces a useable DDPG model for controlling building HVAC systems but fails to induce a useable DQN model. We analyze the reasons behind these observations. Compared to the original room temperature setpoint method, the DDPG-based HVAC control logic achieves a 10.06% energy-saving effect while ensuring comfort.
Optimizing the use of natural ventilation while reducing energy cost_case study for retrofitting an old residential apartment in a Mediterranean city
2023, Energy and Buildings
This study investigates the effects of using a hybrid natural ventilation (NV) and HVAC system in old residential buildings in the Mediterranean city of Beirut. The system operates with a control strategy that maximizes the reliance on NV and reduces the energy cost when the HVAC system is needed. To maximize the performance of the proposed system, this work proposes the addition of insulation panels and the modification of the windows glazing type according to the orientation of the exposed walls. An optimization procedure is carried out to optimize the performance of the insulation layers and the glazing material for each external wall orientation and for the current Mediterranean climate. To this aim, a model to predict the hours of NV, the corresponding penetrating outdoor airflow rate as well as the energy consumption and the average indoor temperature is developed and calibrated using the EnergyPlus software. A genetic algorithm is used to optimize the life cycle cost equation. It was found that depending on the wall orientation, the optimal thermal insulation thickness varied between 20 mm and 40 mm while the glazing U-value varied between 1.7 W/m²·K and 3.2 W/m²·K. These retrofitting strategies and methods resulted in 22% energy savings and 20% more hours of NV per year with respect to the reference case when no retrofit strategies were used.
Optimization-informed rule extraction for HVAC system: A case study of dedicated outdoor air system control in a mixed-humid climate zone
2023, Energy and Buildings
In the era of post-Coronavirus Disease 2019, the dedicated outdoor air system (DOAS), which provides 100% outdoor air for the building, is widely acknowledged as it can ensure acceptable indoor air quality by delivering fresh outdoor air to occupied space. The DOAS with a proper design and operation can provide sufficient ventilation and dehumidification while achieving energy efficiency. Nonetheless, there is limited guidance in determining the optimal control sequence of the DOAS for the designers and operators to implement in practice. Accordingly, in practice, a number of issues have been acknowledged in the design and control phases of DOAS, including insufficient ventilation and dehumidification, and increasing supply air dry-bulb temperature in fear of over-cooling, which might cause significant discomfort and energy waste.
There have been efforts to develop high-performing DOAS controls for better energy efficiency. However, such controls are often complex, or difficult to interpret, for building designers and operators to consider in practice. In this regard, this paper explores a simulation-based framework for generating a supply air temperature control sequence of the DOAS not only to ensure improved energy-saving potential but also to guarantee the implement-ability of the control logic. The U.S. Department of Energy prototype primary school with dynamic occupancy profiles was modeled with a whole building simulation program, EnergyPlus. The model consists of a DOAS with an exhaust air energy recovery system for ventilation and fan-coil units for space cooling and heating. Then, a Genetic Algorithm was adopted to find the true optimal supply air temperature control sequence in terms of minimizing the energy cost of the heating, ventilation, and air conditioning system operation. Lastly, Decision Tree was adopted to extract rules out of the optimums to derive an implementable sequence of operation for the DOAS supply air temperature. A total of 12 week-simulation including four weeks of heating, cooling, and shoulder seasons, separately, under the weather condition of New York City was conducted for the case study. This case study identified that the optimization-informed rule extraction-based control, when compared to conventional outdoor air temperature-based reset control, could save about 13% of energy cost and 25% of energy consumption throughout the heating, cooling, and shoulder seasons. It is notable that the energy-saving was mainly achieved by reducing the heating energy consumption. Importantly, it nearly corresponds to the true optimal control result, which reduces approximately 14% of energy cost and 27% of energy consumption. From the results, it can be highlighted that the optimization-informed rule extraction can be as energy effective as the optimal control, while significantly reducing the complexity of the control.
An adaptive controller based dynamic simulation of household air-conditioner with indirect evaporative cooler as dedicated outdoor air system
2022, Energy and Buildings
The lack of controllable ventilation design for acceptable indoor air quality is a major problem facing small-scale direct expansion air conditioning (DXAC) systems for household applications. The DXAC-IEC hybrid cooling system, consisting of DXAC systems and indirect evaporative cooler (IEC) ventilators, can provide an energy-efficient solution to this problem. An adaptive controller is proposed in this study to respond to changeable indoor and outdoor disturbances to provide standard-compliant fresh air demand, guarantee indoor thermal comfort and reduce energy consumption at the same time. In the adaptive control scheme, different operation modes are applied by combining three modes of compressor speed (high-low, low-off, off) and three modes of IEC fan speed (high, middle, low) responding to outdoor and indoor conditions. The dynamic performance of DXAC-IEC system on typical days and annual performance were numerically investigated. The results show that considerable energy saving (22.1 % to 34.1 % in Xi’an) can be achieved in transition seasons by DXAC-IEC system for providing standard-compliant ventilation rate and guaranteeing indoor thermal comfort simultaneously compared to a standalone DXAC system. However, extra energy (10.0 % to 25.4 % in Xi’an) will be consumed on summer days. On a typical summer day, the indoor relative humidity will increase by 10 % to 15 % because of extra latent load brought from IEC ventilator compared to the standalone DXAC system, but indoor temperature can be precisely controlled and CO₂ concentration is mostly below 1500 ppm.
Performance analysis of small-scale direct expansion air conditioning system with indirect evaporative cooler as dedicated ventilator
2022, Building and Environment
Split-type and window-type air-conditioners are typical small-scale direct expansion air-conditioning (DXAC) system widely-used for household applications. However, lack of controllable ventilation can be a serious problem leading to health problems for such air-conditioning (AC) systems. To improve the indoor air quality of a home space equipped with small-scale DXAC system, an indirect evaporative cooler (IEC) is used as a dedicated ventilator, on one hand, to guarantee proper outdoor air supply and on the other hand, to provide sensible cooling under certain climatic conditions. Validated numerical component models of IEC, DXAC and indoor thermal balance were established and coupled in series to form a closed-loop model of DXAC with IEC ventilator (DXAC-IEC) system. The effects of ambient humidity, ambient temperature, sensible heat ratio and ventilation rate on the system performance were investigated, with comparison to the standalone DXAC system. The results show that the IEC can enhance the system performance under moderate hot and dry ambient conditions (t_amb ≤ 33 °C, RH_amb ≤ 60%) by reducing the cooling load of DXAC. But it would bring extra moisture from outdoor in humid conditions (t_amb ≥ 30 °C, RH_amb>50%) for sacrifice of improved indoor air quality. Optimal fresh air ratio exists for the DXAC-IEC system to achieve the maximum coefficient of performance. The findings will be beneficial for the design and optimization of DXAC-IEC system.
Intelligent control for energy conservation of air conditioning system in manufacturing systems
2021, Energy Reports
Energy conservation is a critical issue, particularly in manufacturing industries with intensive energy usage. The proportional–integral–derivative (PID) controller fails to adjust shop floor temperature against indoor heat dynamics which substantially affect the stability and robustness of the system to achieve energy conservation. By jointly considering the make-up unit (MAU) and a set of collaborative dry cooling coil (DCC) units of a large scaled industrial air condition system, an intelligent control model is proposed in the study to achieve energy conservation and ensure stability of shop floor temperature simultaneously. The temperature set point of the MAU, and the openness of massive distributed DCCs in the factory are optimized in real-time to guarantee the comfort of shop floor temperature, and minimize energy usage and damping. A combined support vector regression with genetic algorithm is proposed to solve the problem efficiently. In addition, the energy efficiency and robustness of the proposed optimization model in securing the stability of the shop floor temperature against indoor heat dynamics is proven through a case study. The total energy consumed by the proposed model is reduced by 7.09% compared with PID controller. The model can regulate shop floor temperature to facilitate operator comfort, product quality, and safe working environment.

View all citing articles on Scopus

View full text

HVAC system optimization with CO2 concentration control using genetic algorithms

Abstract

Introduction

Section snippets

Genetic algorithms

Description of HVAC systems and principle of work

Model of HVAC systems

Simulation results using Matlab Simulink

Result verification using EnergyPlus software

Conclusion

Energy Conversion and Management

Energy and Buildings

Energy and Buildings

Energy and Buildings

Energy and Buildings

Energy and Buildings

Energy and Buildings

HVAC system optimisation—in-building section

Energy and Buildings

HVAC duct system design using genetic algorithms

HVAC&R Research

Building simulation by application of a HVAC system considering the thermal and moisture behaviors of the perimeter walls

Simulation of a centralized cooling plant under different control strategies

HVAC system optimization with CO₂ concentration control using genetic algorithms