Content uploaded by Hanie Omid
Author content
All content in this area was uploaded by Hanie Omid on Jul 22, 2019
Content may be subject to copyright.
Survey of parametric optimization plugins in Rhinoceros used in
contemporary architectural design
Hanie Omid*
M.Arch. student, Pars University of Art and Architecture Tehran, Iran.
Mahmood Golabchi
Professor, School of Architecture, University of Tehran, Iran.
*Corresponding Author’s e-mail: hanie.omid@ut.ac.ir
Abstract: Optimization problems often involve situations in which the user’s goal is to
minimize and/or maximize not a single objective function, but several, usually conflicting,
functions simultaneously. Such situations are formulated as multi-objective optimization
(MOOs) problems. For instance, various building design issues such as building orientation,
indoor thermal comfort, daylighting, life cycle analysis, structural design analysis, energy cost,
etc. can be treated as optimization problems. Recent advances in computational tools for design
applications, coupled with techniques like genetic algorithm (GA) have led to new possibilities
in the way computers can inform and actively interact with the design process. GAs have
attracted much attention in architecture in recent years and are commonly used for solving
MOO problems. Here we review the benefits derived by combining parametric modelling and
energy analysis tools with GAs to achieve a performance oriented process in design, with
specific focus on architectural design. This survey focuses on Grasshopper (GH), a graphical
parametric form generating tool, integrated into Rhinoceros 3D. GH is developed in such a
way to be connected to multiple plugins such as Karamba, LadyBug, Octopus, etc. in order to
facilitate the relation between different disciplines with the same simulation approach. These
plugins are connected to GH and used as an engine to EnergyPlus, Radiance and Daysim for
building energy, comfort, daylighting and lighting simulation as well as carrying out
interactive real simulation, generating form, analyzing structural components and optimization
tasks.
Keywords: Architecture, Parametric design, Optimization, Genetic algorithm, Rhino.
Fourth International Conference on Modern Research in Civil Engineering, Architecture,
Urban Management and Environment, 21 May 2019, Karaj: Iran.
1. Introduction
Interest in tall buildings grew rapidly in the building community around the world during
the past decade as we witnessed the completion of more skyscrapers than any previous decade
in history [1]. Popularity of tall buildings shows their importance in future urbanism and their
efficiency in terms of meeting desired architectural features and structural design requirements.
They, however, exert significant demand on infrastructure and transportation systems, and
affect the historic fabric while reshaping the city’s skyline. They also influence the micro-
environment by casting shadows and blocking views and sunlight, consume massive quantities
of energy and require a high operational cost. Buildings are the biggest single contributor to
world energy consumption, as well as greenhouse gas (GHG) emissions. According to a recent
review [2], energy use in buildings and for building construction represents more than one-
third of global final energy consumption and contributes to nearly one-quarter of GHG
emissions worldwide. Parametric simulation method can be used to improve building
performance [3]. According to this method, the input of each variable is varied to see the effect
on the design objectives while all other variables are kept unchanged. This procedure can be
repeated iteratively with other variables.
Applications of parametric simulation for handling complex engineering systems have
emerged as a promising method [3], and computational optimization to meet the structural
concerns and design considerations should be an integral part of the design process from early
stages of planning. Parametric design of energy efficient buildings using integrated energy
simulation and form optimization with genetic algorithms is an interesting area that shows
promise for improving the architectural synthesis process. Today’s challenges in the buildings
and architectural design are centered around energy, cost, safety, thermal comfort, and carbon
dioxide. To address so many different challenges computational optimization is a must. There
are many instances where multi-objective optimization (MOO) methodology being applied in
the early stages of design to the architectural aspect of tall buildings. MOO process can
improve the workflow efficiency by reducing decision-making time and increasing
collaboration between disciplines [4]. Furthermore, it is a parametric and generative design
methodology which enables the rapid exploration of alternative options and the computational
assessment of trade-offs between environmental performance, energy consumption and cost-
effectiveness. Robust optimization processes could, therefore, significantly reduce carbon-
footprint, reduce the material use and increase the energy-efficiency of the future architecture.
The building architectural components whose performance may be evaluated using the MOO
approach include the façade morphology, envelope material selection, layer thicknesses, etc.
There are a considerable amount of reviews concerning building performance simulations
which provide a comprehensive insight into a specific discipline or branch of building design,
such as: energy simulation, daylighting, optimization, etc. Østergård et al. [5] provided an
overview of the developments in academia and in the software industry related to the use of
building simulations in early building design. Crawley et al. [6] compared 20 building energy
performance simulation programs. Machairas et al. [7] reviewed algorithms for optimization of
building design. Evins [8] reviewed optimization methods applied to sustainable building
design.
Several software packages offer graphical algorithm editors (e.g. Grasshopper), which
are directly linked to 3D modeling tools and allow interactive parametric modeling [9]. The
simplicity of Grasshopper lies within its direct link to one of the most widely used 3D
modeling software Rhinoceros, that controls it with a significant range of tools for
mathematical, geometrical and generative operations, as well as statements of conditionals,
advanced functions and trigonometry. Also the learning curve for its usage is relatively quick
as it does not demand skills of coding or programming of pure scripts – all of its elements are
organized in an easy to understand VPL format. In a parametric design, it is the parameters of a
particular design that are declared, not its shape [9]. That allows designers to assign parametric
values to geometrical objects that can have different spatial arrangement configurations, thus
facilitating the process for form finding. In recent years, digital tools such as Rhino, Revit,
Maya and Ecotect were introduced to the process for simulation, digital form-finding and
fabrication. The rapid development of genetic algorithms (GA) and machine learning tools for
MOO applications, such as Grasshopper (GH) for Rhino and Dynamo for Revit, are bridging
the gap between optimization and visualization, and inter-disciplinary collaboration. According
to Evins [8], the most popular parametric design software is GH, which is a plugin for the
NURBS modeling software Rhinoceros. Furthermore, there are many additional plugins
available for GH to facilitate certain specific and repetitive design tasks, individual developers
design them and very often these plugins are available in an open source format [9].
Here we are concerned with optimization tools in GH. This paper is a survey of GH
plugins used for the building design optimization. For parametric modeling, simulation, form-
finding, energy analysis and optimization of buildings, GH can be coupled with energy
simulation programs such as EnergyPlus, Radiance, and Daysim. This allows the integration of
custom plugins that can then make use of the program’s GUI elements and thus interact with
other plugins in a well-defined way.
2. Methods
Multi-objective optimization (MOO) is a parametric and generative design methodology
which enables the rapid exploration of alternative options and the computational assessment of
trade-offs between environmental performance, energy consumption and capital expense. In
general, a simulation-based MOO method for building optimization procedures consist of six
distinct steps that can be repeated in an iterative manner [2, 4, 7]:
1. Identification of design variables and constraints.
2. Selection of simulation tool and creation of a baseline model.
3. Selection of objective functions.
4. Selection of optimization algorithm.
5. Running simulations until optimization convergence is achieved.
6. Interpretation and presentation of data.
According to a recent review [7], the common objectives to optimize in building
designers, in decreasing order, are energy, cost, thermal comfort, and carbon dioxide. Using
genetic algorithm (GA) solvers such as Octopus and Galapagos plugins for GH, one can
discover the optimal combination of values for a given set of variables Software tools such as
Ladybug and Honeybee (plugins for GH) can be used to connect the parametric geometry to
energy and daylight simulation software (EnergyPlus, Radiance, Daysim, etc.) to support the
decision-making process during the initial stages of design.
2.1 Multi-objective optimization
Optimization is the procedure of finding the minimum or maximum value of a function
by choosing a number of variables subject to a number of constraints [7]. The optimization
function is called cost or fitness or objective function and is usually calculated using simulation
tools. Today, simulation-based optimization has become an efficient measure to satisfy several
stringent requirements of high performance buildings (e.g. increase the energy-efficiency,
daylighting, green buildings, net zero-energy buildings, …). The challenge of the architect is to
create a high performing building design that is the result of often competing objectives.
Optimization methods can be applied to a number of different building design problems such
as massing, orientation, façade design, thermal comfort, daylighting, life cycle analysis,
structural design analysis, energy and cost. When there is more than one objective function for
the optimization problem then a multi-objective optimization (MOO) problem arises. This is
common in building design problems and these functions are often contradictory [7].
In general, MOO requires to implement iterative solution procedure which different
solutions will be investigated by the decision maker. So various combination of the solution
may be needed to reach the optimal solution. The following is a general mathematical
description of the MOO problem [7-8]:
min (f1 (x),f2 (x),…,fk (x)) (1)
s.t.
gi ≤ 0, i=1,2,…,m
hj = 0, j=1,2,…,n
x ∈X
where k≥2 is the number of objectives, fi(x) is the ith object function, x are parameters
and the set X is the feasible set of decision vectors. The feasible set is typically defined by
some constraint functions including: inequality (gi) and or equality (hj) restriction functions.
2.2 Genetic algorithms
Genetic algorithm (GA) is the most common optimization method in the field of building
optimization. In architecture, GA can be adopted in two ways: as an optimization tool or as a
form-generation tool. In the former case, GA deals with well-defined building problems, such
as structural, energy, and daylighting performance. In the latter case, GA is used under the
scope of emergence, i.e., architecture with nature assuming that design is dominated by the
same principles as the natural world. GA uses adaptive heuristics to solve optimization
problems by mimicking the principles of natural selection. First, GA starts with a randomly
generated initial population of potential solutions for the optimization problem. These potential
solutions, often called chromosomes, are coded as binary or real strings. Then, a new
population of children chromosomes is generated from the parent chromosomes through
crossover and mutation procedures). While the crossover process combines parent
chromosomes to produce children chromosomes, the mutation procedure consists of local
modifications of chromosomes. The selection of the chromosomes is achieved based on their
fitness values (Fig. 1). The search of GA is terminated using a convergence threshold within a
tolerable number of generations.
2.3 Optimization tools in Grasshopper
Rhinoceros is a 3D modelling environment, with advanced algorithmic and parametric
modelling capabilities provided by GH. For highly geometrical problems, using a tool with
advanced geometrical capabilities as the basis for an analysis framework saves lots of effort as
many manipulations and analysis functions are provided. There are already a number of
plugins developed for Rhino/Grasshopper as shown in Fig.2. GH allows the integration of
custom plugins that can then make use of the program’s GUI elements and thus interact with
other plug-ins in a well-defined way. For example, structural analysis, energy analysis and
daylight analysis are available from third parties, either as plugins or as easy-to-use interfaces
to standalone software. Some of the favorite plugins are Ladybug, Honeybee, DIVA, and
GECO for environmental analysis; Lunchbox, Kangaroo Physics, and Karamba for structural
analysis; Galapagos, and Octopus for optimization problems. LadyBug/Honeybee are
connected to GH and used as an engine to EnergyPlus, OpenStudio, Ecotect, Radiance and
Daysim for building energy, comfort, daylighting and lighting simulation; DIVA can also
connect to EnergyPlus for energy modeling and it allows the user to parametrically control the
process through GH. The combination of DIVA with other GH plugins like Ladybug or GECO
also allows weather files to be used as inputs, as well as Ecotect analysis results.
Fig. 1. The conceptual procedure of GA
Fig. 2. Various plugins developed for Rhino/Grasshopper
Kangaroo Physics and Karamba are structural analysis plugins in the parametric
environment of GH. LunchBox is used for exploring mathematical shapes, paneling, structures,
and workflow. Kangaroo is used for interactive simulation, form-finding, optimization and
constraint. Karamba makes it simple to combine parameterized complex geometric forms, load
calculations as well as finite element analysis; Galapagos and Octopus are optimization tools
used to find the best solution. Galapagos is plug-in for single objective optimization, whereas
Octopus is for MOO. In particular, GH includes Galapagos which implements GA and
simulated annealing (SA) for single-objective optimization problems. Additionally, there are
four third-party, global optimization plugins including Octopus, Goat, Opossum and Silvereye,
all of which share a similar GUI. Octopus is a Pareto-based optimization tool and implements
the SPEA-2 and HypeE algorithms for MOO; Goat offers an interface with NLopt, a free
library containing various gradient-based and direct search optimization methods; Opossum is
a model-based optimization tool; and Silvereye which implements particle swarm optimization
(PSO) in the GH.
3. Applications of parametric modeling and GAs
Performance criteria in architectural design practice are often multiple and cover several
disciplines, including a building’s geometry, structure and environmental design. This
multidisciplinary motivates studies such as Evins [8], and Gerber and Lin [10]. Nguyen et al.
[2] and Evins [8] surveyed studies of black-box optimization from the perspective of
sustainable building design, while Gerber and Lin [10] presented examples of workflows that
integrate architectural design, performance simulation and black-box optimization based on
custom tools. Hare et al. [11] reviewed studies from structural design. In sustainable building
design, most studies are concerned with optimizing building energy consumption, and in
structural design with reducing structural weight. Other criteria include geometry, daylight and
cost. Banos et al. [12] presented a review of the current state of the art in computational
optimization methods applied to renewable and sustainable energy, offering a clear vision of
the latest research advances in this field. More recently, Eltaweel and Su [9] presented a
literature review on the origin, principle, implementation and applications of parametric design
concept in general, and specifically the relation between parametric design and daylighting and
solar radiation research.
A more specific focus on specific aspects of buildings have been dealt with through
optimization, such as fenestration for daylighting and energy performance by [13, 14]. Looking
at broader aspects, Wang et al. [15] used a multi-objective GA to evaluate design alternatives
for both economical and environmental criteria, while in [16], the authors studied thermal
comfort and energy consumption, concerning design variables affecting passive solar behavior
such as size of the windows and thermal mass and HVAC systems. Gagne and Andersen [17]
focused on the use of MOO in the optimization of facade morphologies for daylight
penetration and glare control using SketchUp 3D modeling software. In two papers [18-19],
GH were used to link energy simulation programs, EnergyPlus in both cases, with optimization
algorithms and conduct design optimization. Chang and Shih [20] studied the design of a
public apartment in Taiwan, for which the environmental impact, sustainability and
construction cost efficiency have been set as the objectives in the planning phase. GA was used
for the optimization-simulation, while a detail whole building simulation program, is used for
lighting, energy and cost analysis. They adopted DIVA for lighting and energy simulation,
which allows the modeling of single-zone thermal models using EnergyPlus and lighting using
Daysim, and optimization procedure combined Galapagos plugin and scaled sum of weighted
objective functions.
Waibel and Evins [21] applied a geometrical optimization process to an urban district of
12 buildings, using detailed energy simulations to evaluate their performance. The objectives
were total floor area, lighting energy demand, and space conditioning energy demand. The
geometric modelling, energy simulation and optimization framework employed in this work is
based on the Rhinoceros / Grasshopper platform and associated plugin Archsim as a link to
EnergyPlus; Octopus for MOO. Qingsong and Fukuda [22] used parametric design in their
study to optimize daylight and heat gain inside an office building in Beijing. The study method
depends on finding the best size window area on each wall to minimize the energy
consumption and maximize the useful daylight illuminance. The experiment done by creating
the model in GH, and the climatic simulation was done using Ladybug and Honeybee as an
engine to EnergyPlus, Radiance and Daysim, in order to study the environmental state
parametrically. Additionally, the study used Galapagos plugin to achieve the role of
optimization, which provides a generic platform for the application of evolutionary algorithms
to be used on a wide variety of problems by non-programmers. In one of their case studies,
they introduced Grasshopper and its plugins, ladybug, honeybee and Galapagos, with a
simplified tool for building thermal evaluation for the purpose of minimizing the energy
consumption and maximize the useful daylight illuminance of an office building in China
Beijing. Touloupaki and Theodosiou [23] proposed a nearly Zero Energy Buildings (nZEBs)
design workflow, integrating evolutionary algorithms and energy simulation. The proposed
workflow methodology combines Parametric modelling and MOO algorithms were used to
integrate energy simulation in the early design stages of a building, in order to minimize its
lifecycle energy requirements and achieve the nZEB standards. The software tools proposed
for seamless operation were GH with Galapagos evolutionary solver, Ladybug and Honeybee
plugins. Ladybug and Honeybee were used to connect the parametric geometry to energy and
daylight simulation software (EnergyPlus, Radiance, Daysim, etc.) to support the decision-
making process during the initial stages of design.
Agirbas [24] developed a model capable of optimizing the minimal surface based form as
the roof cover of the building according to the determined objectives, constraints and variables.
The optimization of the form was based on daylighting, radiation and covered area parameters.
In this context, using a GA based MOO tool, a model was developed whose objectives are
reducing radiation, increasing daylight in the space and increasing the floor surface area of the
form. During this process, GH was used as a visual programming language platform and the
Octopus plug-in was used as a MOO tool (Fig. 3). In addition, Honeybee plugin was used to
connect GH to simulation engines (EnergyPlus, Radiance, Daysim and OpenStudio) for
daylight analysis. Ladybug plugin was used import and analyze standard weather data for
radiation analysis. MinSurf plugin was used for creating minimal surface geometry. Flow chart
of the proposed methodology is shown in Fig. 3.
4. Conclusion
Computational design is where the architecture and construction industries benefit from
the technological advancements in the software industry to develop more efficient designs.
Modern design and planning tools provide an easy way to harness the power of computation in
a design process without having to learn how to write code. These tools allow architects and
designers create their own tools, and let a reunion of the disciplines. Effective optimization of
building optimization problem involves coupling a building energy simulation program with an
optimization evolutionary algorithm such as the genetic algorithm. Such a multi-objective
optimization (MOO) process in the early stages of design can improve the workflow efficiency
by reducing decision-making time and increasing collaboration between disciplines. In this
way, the design process simultaneously combines architecture and engineering.
The most popular software for computational performance-driven design optimization
among architects is GH for Rhinoceros 3D. A possible explanation for this trend is that Rhino
is a widely-used software among architects and GH is a free tool with many developers that
constantly provide support and new components. Moreover, GH features a multitude of already
developed tools and plugins, such as Ladybug, Honeybee, Kangaroo, Galapagos, Octopus, and
others, that provide many capabilities and render it an accessible method of using
environmental performance simulation via a visual programming interface. Software such as
GH could well fill the gap between the design and construction, and make a connection
between the building industry and architectural design.
Fig. 3. Software platforms for minimal surface based form optimization [24]
References
[1] Ali MA, Al-Kodmany K. (2012). Tall buildings and urban habitat of the 21st century: a
global perspective. Buildings, vol. 2, p. 384-423.
[2] Omid H, Golabchi M. (2018). A review on green building architecture utilizing
environmentally friendly energy-saving technologies. 5th National Conf. & 1st Int'l
Conf. on Research in Civil Engineering, Tehran – Iran, March 10.2018.
[3] Nguyen AT, Reiter S, Rigo P. (2014). A review on simulation-based optimization
methods applied to building performance analysis. Appl Energy, vol. 113, p. 1043–58.
[4] Shen X, Singhvi A, Mengual A, Spastri M, Watson V. (2018). Evaluating the multi-
objective optimization methodology for performance-based building design in
professional practice. ASHRAE and IBPSA, Chicago, USA, p. 646–653.
[5] Østergård T, Jensen RL, Maagaard SE. (2016). Building simulations supporting decision
making in early design – a review. Renew Sustain Energy Rev, vol. 61, p. 187–201.
[6] Crawley DB, Hand JW, Kummert M, Griffith BT. (2008). Contrasting the capabilities of
building energy performance simulation programs. Build Environ, vol. 43, no. 4, p. 661–
673.
[7] Machairas V, Tsangrassoulis A, Axarli K. (2014). Algorithms for optimization of
building design: a review. Renew Sustain Energy Rev, vol. 31, no. 1364, p. 101–112.
[8] Evins R. (2013). A review of computational optimization methods applied to sustainable
building design. Renew Sustain Energy Rev, vol. 22, p. 230–245.
[9] Riekstins A.(2017). Advanced modeling techniques for architectural design education ,
Architecture and Urban Planning, vol. 13, no. 1, p. 106-111.
[10] Gerber DJ, Lin SHE. (2014). Designing in complexity: Simulation, integration, and
multidisciplinary design optimization for architecture. Simulation, vol. 90, p. 936–959.
[11] Hare W, Nutini J, Tesfamariam S. (2013). A survey of non-gradient optimization
methods in structural engineering. Adv Eng Softw, vol. 59, p. 19–28.
[12] Banos R, Manzano-Agugliaro F, Montoya FG, Gil C, Alcayde A, Gómez J. (2011).
Optimization methods applied to renewable and sustainable energy: a review. Renew
Sustain Energy Rev, vol. 15, no. 4, p. 1753–1766.
[13] Caldas L, Norford L. (2002). A design optimization tool based on a genetic algorithm.
Automation in Construction, vol, 11, p. 173–184.
[14] Wright J, Mourshed M. (2009). Geometric optimization of fenestration. Proceedings of
the Eleventh International IBPSA Conference, Glasgow-UK, July 27–30.2009, p. 920–
927.
[15] Wang W, Zmeureanu R, Rivard H. (2005). Applying multi-objective genetic algorithms
in green building design optimization, Building and Environ, vol. 40, p. 1512–1525.
[16] Magnier L, Haghighat F. (2010). Multiobjective optimization of building design using
TRNSYS simulations, genetic algorithm, and artificial neural network, Building and
Environ, vol. 45, no. 3, p. 739–746.
[17] Gagne JML, Andersen M. (2010). Multi-Objective façade optimization for daylighting
design using a genetic algorithm, SimBuild, 4th National Conference of IBPSA, New
York-USA, 2010.
[18] Glassman E, Reinhart C. (2013). Façade optimization using parametric design and future
climate scenarios. Proceedings of the 13th International IBPSA Conference, 2013.
[19] Hollberg A, Ruth J. (2014). A parametric life cycle assessment model for façade
optimization. Proceedings of the 2nd BSO Conference, 2014.
[20] Chang MC, Shih SG. (2015). A hybrid approach of dynamic programming and genetic
algorithm for multi-criteria optimization on sustainable architecture design, Comput
Aided Des Applic, vol. 12, p. 310–319.
[21] Waibel C, Evins R. (2015) Exploring the use of variable mapping for optimizing urban
morphologies, Proceedings of BS2015, Hyderabad-India, Dec. 7-9.2015.
[22] Qingsong M, Fukuda H. (2016). Parametric office building for daylight and energy
analysis in the early design stages. Procedia Soc. Behav. Sci. vol. 216, p. 818–828.
[23] Touloupaki E, Theodosiou T. (2017). Optimization of building form to minimize energy
consumption through parametric modelling. Procedia Soc. Environ. Sci. vol. 38, p. 509–
514.
[24] Agirbas A. (2018). Performance-based design optimization for minimal surface based
form. Architectural Sci Rev, vol. 61, no. 6, p. 384-399.
Fourth International Conference on Modern Research in Civil Engineering, Architecture,
Urban Management and Environment, 21 May 2019, Karaj: Iran.