One of the main problems in the AEC industry is HVAC consumption in buildings. The reasons for reducing energy consumption vary. There are clear economic reasons, as a good energy consumption strategy could lead to reduced energy costs and impact financial savings for consumers. Reducing energy use is also a solution to reduce greenhouse gas emissions. This work implements a Micro-Genetic Algorithm (MGA) to reduce heating and air-conditioning consumption by modifying a simple building's geometry (a house) and the position and size of two windows.
The results showed that after few iterations the MGA was able to reduce the HVAC consumption in around 17%.
Description
A big window
The sun shining through a big window in summer will heat the furniture and air, increasing the temperature in the room. Therefore, the air conditioning system will have to work to keep the room at 20ºC. The positive side is that we can take advantage of natural light for a long time. We can also cool the house by opening the window and releasing the heat trapped in the walls and furniture through radiation during the day.
In winter, the same window will capture heat, but unfortunately, all the heat will be lost very quickly as it escapes through the same window where it entered. Therefore, we will have to turn on the heating quickly. The problem is that all the heating produced will continue to escape through the window, increasing heating consumption.
A small window
small window in summer will not help release the heat trapped in the walls. The air conditioning will run for more hours to keep the temperature at 20ºC, and it will not be possible to take advantage of natural light as much as with the big window.
In winter, the same small window will help retain the temperature in the room.
A big window performs well in summer but has lower performance in winter. On the contrary, the small window is not as useful in summer but performs better in winter compared to the big window. The solution does not consist of simply making a big or small window but finding the optimum balance in both windows' dimensions and positions.
The objective is to answer the following question:
What are the building’s geometry and the windows’ dimensions and positions to keep the temperature at 20ºC during the hottest and coldest days of the year with the lowest energy consumption?
Micro Genetic Algorithm (MGA)
An MGA is similar in architecture to a Genetic Algorithm. The MGA uses small populations and few generations (15 in this work) in comparison to normal GAs, where the population size and number of generations could easily reach the thousands. The MGA implements different strategies to avoid fast convergence and poor solutions. The strategy consists of "locating" the highest-ranked individual in the population in a promising area of the search space (elitism). A pool of individuals is used to control diversity during the generations. This technique also involves generating random individuals. The idea behind it is to explore the search space further and avoid finding only local minima.
Micro Genetic Algorithms show several benefits. For example, since the algorithm is not evolving large populations, convergence can be achieved quickly, and less memory is required to store populations. In some cases, the fitness function evaluation could take minutes or hours, making it impractical to work with large populations. This point is very interesting for the AEC industry because energy consumption calculations in large buildings are time-consuming. Another reason to implement an MGA is its use in high-dimensional problems. The individuals in the population recall large sizes, which implies a large number of cost-function evaluations.
The MGA was implemented in laga and generative components was used as parametric software. Energy plus software was used for energy modeling and fitness evaluation. The connection between Generative Components and Energy Plus was implemented in C# and GC script.
The MGA in action.
Parametric model
The parametric model is composed of 6 flat surfaces and 2 windows, facing East and West. The box (house) is 12m in length (X direction), 10m in width (Y direction), and 4m in height. The box-house geometry is controlled by 4 parameters. The windows are placed on planes hosted on the East and West walls. Several vectors are used to define the dimensions and positions of the windows. Once the model was built, a test with many iterations was conducted to detect any inconsistency in the parametric definition.
Energy model
The energy model is essential because it sets the physical characteristics of the box-house, such as the building’s location, weather conditions, wall insulation and thermal properties, window properties, house geometry, window dimensions, and positions. This information is defined in an Energy Plus file.
Energy model parameters: *Building location*: Country terrain, Chicago IL. Windows: Double Panel window with an interior camera of 3 mm. Walls: Wood siding outside, fiberglass quilt in the middle, and plasterboard in the interior. Roof: Roof deck outside, fiberglass quilt, in the middle and plasterboard in the interior. Floor: Thickness(m): 0.10, Conductivity(W/m-K): 1.7296, Density(kg/m3): 2243.0, Thermal Absorptance: 0.9, Solar Absorptance: 0.65, Visible Absorptance: 0.65 HVAC system**: Heating Supply Air Temperature(C): 50, Cooling Supply Air Temperature(C): 13, Heating Supply Air Humidity Ratio(kg-H2O/kg-air): 0.015, Cooling Supply Air Humidity Ratio(kg-H2O/kg-air): 0.01
The analysis correspond to the consumption of heating and cooling during the hottest and coldest day of a year.
Workflow
The code transforms the polygon vertices in the parametric model into a list of strings. After that, it analyzes the Energy Plus file (IDF file format) to create a new IDF with the recently modified vertices (new house-box geometry) and windows (size and position). Once the new IDF file is created, the code calls Energy Plus to execute the energy analysis and waits until the energy report is finished. The code reads the report and extracts heating, cooling, and room temperature data. This data is used as a fitness measure to classify the different individuals in the population.
The individuals are composed of a list of 31 doubles. The first 30 values represent the X, Y, and Z coordinates of the parametric model, and the last value is the fitness function. Once the individual is created, it is saved into the array population. After the population is created, the elitism selection is executed.
The next step in the MGA is the crossover, in which the highest-ranked individuals exchange parts of their chromosomes to create a new candidate solution (offspring), inheriting the properties from their parents. The crossover is performed only in certain parts of the chromosome, while the rest is set randomly to avoid local minima. Since the population is small and the algorithm was only tested over 5 generations, it is necessary to explore all possible solutions during the short period in which the MGA runs.
The offspring's genes are mutated randomly by a factor. In this work, the best-tested results set this value around 0.5. The mutation is calculated using a random value between 0 and 1, which is then added or subtracted to one of the values in the gene:
offspringMutate[i] = low + r * (low – high) (2
The mutated offspring and the highest-ranked solution from the previous generation are used in the new population, with 3 new individuals randomly generated to load the new population.
Results
The MGA was tested 14 times with no variation in the data environment, energy properties, or number of generations. The results were as follows: In 6 test, the MGA was able to optimize in each 5 generation In 6 test, the MGA was able to optimize in 4 generations In 1 test, the MGA only optimize in 1 generation In 1 test, the MGA was unable to optimize.
The energy consumption average at the beginning of the 12 successful optimization was: 4,100,503 Jules means: 1139.029 W/h. The energy consumption average after the optimization was: 3,494,576 Jules means: 970.715 W/h. This represents a reduction of 17.21% in energy consumption during the hottest and coldest day of the year. The best individual in all successfully tests reduced the energy consumption to: 3,023,091.22 Jules means: 840.412 W/h.
The energy consumption average at the beginning of the 12 successful optimizations was 4,100,503 Joules, which means 1139.029 W/h. The energy consumption average after the optimization was 3,494,576 Joules, which means 970.715 W/h. This represents a reduction of 17.21% in energy consumption during the hottest and coldest days of the year. The best individual in all successful tests reduced the energy consumption to 3,023,091.22 Joules, which means 840.412 W/h.
The above images show the 12 successful tests with the energy consumptions before and after the optimization process.
Conclusion
The results show the MGA performed, in most cases, very well in decreasing HVAC consumption by around 17.00%. The MGA "bent" the box-house to reduce the area exposed to the East and West and also reduced the total volume of the interior space, which had a direct impact on HVAC consumption.
The windows' size and position were set randomly to avoid local minimums. Nevertheless, it is possible to observe similarities in window size and position in the successful tests. For example, the MGA preferred small windows located near the top wall edges. The East and West windows were always located on the top right side of the wall. The MGA also proposed bigger, vertically shaped East windows. On the other hand, the West windows were smaller and horizontally shaped.
The MGA located the West horizontal window near the top edge of the wall 10 times. In contrast, the East window was generated near the bottom wall edge 6 times and near the top edge 4 times. In all cases, the MGA tried to reduce the total volume of the building and bend the box-house geometry, with East and West surfaces shifted to the South 10 times and to the North 2 times.
The MGA was able to generate 13 possible solutions with low energy consumption levels out of 14 tests. This makes the box-house more sustainable, which is an issue of concern for the AEC industry and is also reflected in the monthly building budget.
The minimum window size was not constrained, which can result in an uncomfortable spatial relationship between the interior and the exterior of the building. Therefore, this parameter should be revised in the following steps. This box-house energy analysis takes about 3.7 seconds per evaluation, including the necessary report to read the solutions and inform the MGA of the HVAC consumption.
One of the main problems with small populations is the inability to explore the search space efficiently, often only finding local minimums. The strong inheritance in a complete crossover stops the optimization in the second or third generation, falling directly into local minimums.
A much more complex geometry, optimized by a Micro Genetic Algorithm.
