You are on page 1of 8

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/238521415

Design of a toy submarine using underwater vehicle design optimization


framework

Article · April 2011


DOI: 10.1109/CIVTS.2011.5949527

CITATIONS READS
9 1,527

3 authors:

Khairul Alam Tapabrata Ray


UNSW Sydney UNSW Sydney
11 PUBLICATIONS   257 CITATIONS    305 PUBLICATIONS   6,763 CITATIONS   

SEE PROFILE SEE PROFILE

S.G. Anavatti
Australian Defence Force Academy
288 PUBLICATIONS   3,292 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Novel Constraint Handling Strategies for Computationally Expensive Optimization Problems View project

Many Objective Computationally Expensive Constrained Optimization View project

All content following this page was uploaded by S.G. Anavatti on 01 September 2014.

The user has requested enhancement of the downloaded file.


Design of a Toy Submarine Using Underwater
Vehicle Design Optimization Framework
Khairul Alam, Tapabrata Ray and Sreenatha G. Anavatti
School of Engineering and Information Technology
University of New South Wales
Australian Defence Force Academy
UNSW@ADFA, Canberra, ACT 2600, Australia
Telephone: +61 2 6268 8479, Fax: +61 2 6268 8276
Email: khairul.alam@student.adfa.edu.au, t.ray@adfa.edu.au, s.anavatti@adfa.edu.au

Abstract—This paper presents a framework for optimum gliding down in depth and using buoyancy to surface), Re-
design of a small, low-cost, light-weight toy submarine for recre- motely Operated Vehicles (ROVs), Autonomous Surface Ve-
ational purposes. Two state of the art optimization algorithms hicles (ASVs) and Autonomous Underwater Vehicles (AUVs).
namely Non-dominated sorting genetic algorithm (NSGA-II) and
Infeasibility driven evolutionary algorithm (IDEA) have been In particular, the UUVs have been promoted as a prominent
used in this study to carry out optimization of the toy submarine option and tool in oceanographic exploration and exploitation
design. Single objective formulation of the toy submarine design programs.
problem is considered in this paper to identify designs with Traditionally, the UUV design process has been largely
minimum drag and internal clash free assembly. The flexibility ‘ad hoc’ with designs governed by experience and rules of
of the proposed framework and its ability to identify optimum
preliminary designs of a toy submarine are demonstrated. Design thumb [3]. The application of formal optimization methods
identified through the process of optimization is compared with to the drag minimization or to evaluate optimum design of
an existing toy submarine to highlight the benefits offered by the UUVs have not gained much attention by the researchers so
present approach. far [4]. It is important to highlight that the use of efficient
optimization tools leads to better product quality and improved
N OMENCLATURE functionality [5].
D Drag In the recent years, some works have considered the prob-
d Maximum body diameter lem of finding the optimum hull form for UUVs which include
dt Smaller diameter of the tail the works of [3], [6]–[8] to minimize drag. Small improvement
l Length overall in drag can result in a substantial saving in thrust requirement.
lm Length of the parallel middle body However, much work still needs to be done in terms of
ln Length of the nose optimizing the hull form design to minimize drag and increase
lt Length of the tail propulsion efficiency [9].
LA1 Length of the first lever arm Although a number of present works have been devoted to
LA2 Length of the second lever arm find the optimum hull form for UUVs, limited attention has
nn Shape variation coefficient of the nose been paid towards the optimum ‘clash-free’ placement of the
nt Shape variation coefficient of the tail internal components and factors affecting controllability, i.e.
s Longitudinal distance between CB and CG the centre of gravity (CG) and centre of buoyancy (CB) effects.
ZB Z-coordinate of the battery unit This paper presents a framework for design optimization of a
ZC Z-coordinate of the controller toy submarine that represents a class of torpedo-shaped un-
ZL Z-coordinate of the propeller unit for lateral movement derwater vehicles by simultaneously considering both internal
ZV Z-coordinate of the propeller unit for vertical move- clash-free arrangement of on-board components and external
ment size and shape for given design requirements. The objective
is to find an appropriate hull shape to minimize drag and
I. I NTRODUCTION optimum clash-free placement of the internal objects for opti-
mal CG/CB separation thereby ensuring better controllability
Unmanned underwater vehicles (UUVs) have received of a vehicle moving submerged near the free surface while
worldwide attention and been widely used in ocean explo- fulfilling the design constraints.
ration, military and industrial applications [1], [2]. The wide The organization of this paper is as follows. Section II
range of applications have resulted in development of hundreds provides a framework for the design optimization of the toy
of UUVs with a variety of shapes, sizes, working depth limits, submarine. Thereafter, the details of the numerical experi-
sources of energy, means of propulsion and ways of control. ment are described in Section III. The numerical results are
Such vehicles include glider (unpropelled underwater vehicles presented in Section IV. Finally, conclusions are made in

978-1-4244-9976-2/11/$26.00 ©2011 IEEE


Section V. 1) Hull Geometry and Material Selection: The hull size
of the toy submarine is constrained by the space for the on-
II. O PTIMIZATION FRAMEWORK
board instruments that needs to be carried, and the hull shape
This work presents an optimization framework for the is constrained by the hydrodynamic characteristics for min-
design of a toy submarine with minimum drag considera- imization of drag. Axisymmetric body of revolution moving
tion. The framework incorporates geometry and configuration submerged near to the free surface is considered in this study.
modules, a hydrodynamics module, several accepted maritime Illustration of the parameterization for the hull geometry is
performance and characteristics estimation methods and a suite shown in Fig. 2, where the body is seen to comprise a nose-
of optimization algorithms. Shown in Fig. 1 is the flowchart section of variable length ln , a mid-section of variable length
of the optimization framework. The framework facilitates the lm , and a tail-section of variable length lt , making up the total
communication of data from one application to the next, body length of l units.
producing an automated multidisciplinary design environment.

Fig. 2: Parameterization of the hull geometry

The nose radius follow from the Eq. (1),


  n  1
1 ln − xn n nn
rn = d 1 − (1)
2 ln
where d is the maximum body diameter in m, which may be
varied, and nn is the shape variation coefficient of the nose
which may also be varied to give different shapes, as shown
in Fig. 3.

Fig. 1: Detail flowchart of the optimization framework

A. Geometry and Configuration Modules


The basic design concept of the toy submarine employed Fig. 3: Parameterization of the nose geometry
herein is based on the development of a small, light-weight
vehicle that can be easily launched, recovered and operated For the capability of housing the rear propeller and ease of
without any special handling equipments capable of working fabrication a long tapered tail cone is used in this study. This
at snorkeling depth. The design requirements chosen in this shape follows the fluid flow lines and as the contraction is still
study are: occurring at the nozzle exit, the velocity of the fluid is still
• Speed is 0.5 m/s increasing as it leaves the nozzle.
• No longer than 400 mm The materials for the design vehicle should have the fol-
• Total weight be less than or equal to 450 g lowing properties:
• The vehicle to be propelled by one rear propeller and two • good resistance to corrosion
propellers each for vertical and lateral movements • high strength but relatively lightweight
• Cost effective • ease of fabrication: can the vehicle be manufactured
Considering the basic design requirements, the geometry easily with the chosen material?
and system configurations for the toy submarine are formu- • cheap
lated as follows. • availability
parts of the toy submarine for optimal CG position. The
term clash-free, is defined as the placement of the internal
components such as controller, propellers and battery compart-
ment in their respective positions without any overlapping as
well as keeping appropriate clearance among them. Practically,
the components considered here are irregular in shape. How-
ever, for simplicity of the clash-free mechanism, we considered
minimum bounding box dimensions of those components in
Fig. 4: Parameterization of the tail geometry this study.
Once the internal parts are placed in a clash free state, the
parallel middle body is generated automatically covering the
The Polyvinyl chloride (PVC) water pipe for it’s intrinsic internal arrangement, and then nose, rear propeller and tail
properties fits best for all the above requirements, and there- cone are attached along with the mid-body, thereby generating
fore, has been chosen for this study. the whole toy submarine shape. In the present study, Matlab is
2) Propulsion System: The propulsion of the toy submarine used for geometry manipulation, and interfacing with CATIA
under consideration is achieved through the use of three is done using VBScript macros for the vehicle design within
propeller units, one of which is used for vertical movement, this framework.
i.e. to move the vehicle up and down. Another propeller unit B. Hydrodynamics Module: Drag Estimation
is used to steer the vehicle left and right, and the remaining
propeller is used to propel the vehicle forward and backward The hull for the design vehicle has been optimized for
in the water, as illustrated in Fig. 5. minimum drag that eventually reduces the submerged vehicle
power requirements [4]. For drag estimation, the following
formula has been used:
1 2
D= ρV CV S (2)
2
where ρ is the density of the fluid in kg/m3 , V is the velocity
in m/s, S is the wetted surface area of the vehicle in m2 , and
CV is the coefficient of viscous resistance for the smooth bare
hull.
Three methods- Virginia Tech (VT) [3], MIT [13] and
G&J method [14] are employed in this study to measure the
Fig. 5: Configuration of the propulsion system coefficient of viscous resistance (CV ) in three different ways
to ensure uniformity in drag estimation of the design vehicle.
To achieve the required design speed of 0.5 m/s, optimiza-
tion of the entire vehicle needs to be done to reduce drag to C. Optimization Modules
a minimum while increasing thrust to a maximum to improve In this framework, two state of the art optimization al-
the efficiency of the propellers. Therefore, the position of the gorithms, NSGA-II and IDEA are used. The algorithms are
propeller units for vertical and lateral movements as illustrated written in Matlab and are integrated with CATIA along with
in Fig. 5, are not fixed, but are rather free to move within the VBScript to automate the whole design process.
entire mid-section. During optimization process, the optimizer • NSGA-II: Non-dominated Sorting Genetic Algorithm
chooses the optimal positions of those propeller units while II (NSGA-II) proposed by Deb et al. [15], is one of the
designing the vehicle. most popular population based optimization algorithms
3) Power Source: Most underwater vehicles in use today that has been successfully used in a number of real life
are powered by low cost rechargeable batteries. A few of the applications. The main steps of NSGA-II are outlined in
larger vehicles are powered by solid Polymer Electrolyte Fuel Algorithm 1. It uses Simulated Binary Crossover (SBX)
Cell (PEFC) [10], aluminium/oxygen fuel cell [11] and by and polynomial mutation for generating off-springs. Indi-
solar energy [12] where endurance is a keen issue. However, vidual solutions in a population are ranked based on their
they require substantial maintenance and expensive refills for fitness value. Feasible solutions are considered better than
continuous deployment. Due to simplicity and commercial infeasible solutions and are ranked higher. Feasible and
availability, the AA size alkaline batteries with nominal volt- infeasible solutions are ranked separately. For single ob-
age of 1.5 V and energy density of 140 Whr/kg have been jective optimization feasible solutions are sorted based on
selected in this study. These batteries are attractive as they are the objective value while for multi-objective optimization
cheap, small and usually safe. the solutions are ranked based on non-dominance.
4) Clash-free Mechanism: A remarkable point of this work • IDEA: Solutions to real-life constrained optimization
is to apply ‘clash-free’ mechanism while arranging the internal problems often lie on constraint boundaries. In reality,
Algorithm 1 Non-dominated Sorting Genetic Algorithm III. N UMERICAL E XPERIMENT
II (NSGA-II)
Require: NG > 1 {Number of Generations} In this section, the optimization problem formulation is
1: Initialize (P1 ) {Create an initial population of solutions} presented. Based on the design requirements, the following
2: Evaluate (P1 ) problem has been formulated.
3: Rank (P1 ) {Assign ranks to each solution}
4: for i = 2 to NG do Single objective optimization
5: Ci = Evolve (Pi−1 ) {Create child solutions from parents The single objective optimization problem is posed as the
of previous generation} identification of a vehicle hull form with minimum drag as
6: Evaluate Ci {Compute the performance of the child well as ‘clash-free’ optimal placement of the internal objects
solutions} subject to the constraints on lever arms and CG/CB separation.
7: Rank (Pi−1 + Ci ) {Assign ranks to each solution} The first (LA1 ) and second (LA2 ) lever arms are the longitu-
8: Pi = Reduce (Pi−1 + Ci ) {Identify parents for the next dinal distances of the propellers from the centre of buoyancy
generation} respectively. The higher value of lever arm produces higher
9: end for pitching and turning moments that lead to better diving and
heading changes. The CG/CB separation (s) is the longitudinal
distance of the centre of gravity from centre of buoyancy. The
lower the value of s, the closer the position of the CG and
a designer is interested to look at the solution that
CB that leads to better stability of the vehicle. Minimization of
might be marginally infeasible. NSGA-II and most other
drag is important because minimum drag leads to least power
optimization algorithms intrinsically prefer a feasible
consumption for propulsion, and corresponding savings in the
solution over an infeasible solution during its course of
operating costs. The objective function, constraints and design
search. However, Infeasibility Driven Evolutionary Algo-
variables are listed in the Eq. (3).
rithm (IDEA) introduced by Singh et al. [16], suggests
that preservation of marginally infeasible solutions during
the course of search can expedite the rate of convergence. Minimize:
Since the Pareto optimal solutions for a constrained f (1) = D
problem usually lie on a constraint boundary, IDEA tries
to focus the search near the constraint boundaries by Constraints:
maintaining a set of infeasible solutions (in addition to g (1) = LA1 ≥ 45 mm; g (2) = LA2 ≥ 90 mm
feasible solutions) near the constraint boundary during g (3) = s ≤ 4 mm
the search. The main steps of IDEA are outlined in (3)
Algorithm 2. IDEA ranks the infeasible solutions higher Design variables:
than the feasible solutions to provide a selection pressure 0 ≤ ZC ≤ 300 mm; 0 ≤ ZV ≤ 300 mm
to create better infeasible solutions resulting in an active 0 ≤ ZB ≤ 300 mm; 0 ≤ ZL ≤ 300 mm
search through the infeasible search space. Interested 35 ≤ dt ≤ 50 mm; 80 ≤ lt ≤ 150 mm
readers are referred to Ray et al. [17] for a comparison of 1.5 ≤ nn ≤ 3; 45 ≤ ln ≤ 100 mm
IDEA with other state of the art optimization algorithms.
The above formulated problem is solved using NSGA-II
Algorithm 2 Infeasibility Driven Evolutionary Algorithm and IDEA. Multiple runs are performed varying parameters-
(IDEA) crossover probability, mutation probability, crossover distri-
Require: N {Population Size} bution index and mutation distribution index. The values of
Require: NG > 1 {Number of Generations} the parameters are listed in Table I, and for each parameter
Require: 0 < α < 1 {Proportion of infeasible solutions} combination, both the algorithms are run with two different
1: Ninf = α ∗ N
random seeds.
2: Nf = N − Ninf
3: pop1 = Initialize() TABLE I: Parameters used for NSGA-II and IDEA
4: Evaluate(pop1 ) Parameter Value
5: for i = 2 to NG do Crossover probability 0.9
Mutation probability 0.1
6: childpopi−1 = Evolve(popi−1 ) Crossover distribution index 10
7: Evaluate(childpopi−1 ) Mutation distribution index 20
8: (Sf , Sinf ) = Split(popi−1 + childpopi−1 )
9: Rank(Sf ) A population size of 500 is evolved for 300 generations,
10: Rank(Sinf ) resulting in 150,000 function evaluations for each run. Twenty
11: popi = Sinf (1, Ninf ) + Sf (1, Nf ) independent runs are done using NSGA-II and IDEA. For
12: end for
IDEA, infeasibility ratio (α) of 0.2 is used.
0.22
IV. R ESULTS AND D ISCUSSION NSGA−II
IDEA
A. Single Objective Optimization Results 0.2

Shown in Fig. 6 is the result of the best run for the


0.18
single objective drag minimization problem using NSGA-II
and IDEA. 0.16

Drag (N)
0.16 0.14
NSGA−II
IDEA
0.15
0.12
0.14
0.1
0.13
Drag (N)

0.08
0.12 0 5 10 15
Function evaluations x 10
4

0.11
Fig. 7: Progress plot of median design for single objective
0.1 drag minimization using NSGA-II and IDEA
0.09

0.08
0 5 10 15
Function evaluations x 10
4

Fig. 6: Progress plot of the best design for single objective


drag minimization using NSGA-II and IDEA Fig. 8: CATIA model of the resulting optimized toy submarine

It can be observed that both optimization algorithms are Propeller unit for Propeller unit for
Tail
Nose vertical movement lateral movement
able to achieve near optimal value of drag in approximately
37000 function evaluations. The statistics of results computed
across 20 runs for each algorithm is reported in Table II. It
is seen that the best, median and the average objective values
obtained by IDEA are better than NSGA-II. In addition, the
Controller unit Battery compartment Propeller unit for
standard deviation across the multiple runs is much less than longitudinal movement
NSGA-II, indicating that it is able to achieve better objective
values more consistently. This is also reflected in the median Fig. 9: Configuration of the resulting optimized toy submarine
runs as shown in Fig. 7, where IDEA is seen to converge faster
than NSGA-II.

TABLE II: Single objective drag minimization results


Design NSGA-II IDEA
Best 0.084003 0.082177
Mean 0.086081 0.083828
Median 0.0859657 0.0841192
Worst 0.087465 0.086806 Fig. 10: Longitudinal section of the resulting optimized toy
SD 0.000937534 0.000892474 submarine

B. Results of Optimum Toy Submarine Design submarine has independent propellers to allow it to ascend,
Based on the results obtained by carrying out optimization descend, turn and move forward and backward. Configuration
of drag, Figs. 8-10 show the optimal shape and internal of the internal components of the toy submarine is shown in
configurations of the optimized toy submarine. Fig. 12. The specifications and performance criteria measured
The resulting performance criteria of the optimized toy sub- for this toy submarine are also listed in Table III.
marine are listed in Table III. The values of the lever arms and It can be observed that the performance criteria of the
CG/CB separation as reported in Table III, clearly indicate that resulting optimized toy submarine are close to the results
the design constraints are satisfied while achieving minimum measured for the existing toy submarine. In addition, the
drag. length of the first lever arm of the optimized toy submarine
An example of similar existing toy submarine available in is higher and also the value of CG/CB separation is lower
the market is USS Dallas RC toy submarine. This model than that of the existing toy submarine; thereby ensuring better
efficient integrated analysis/design system comprising high-
end mathematical (Matlab) and CAD package (CATIA), the
framework is able to generate an optimized geometry of
the toy submarine based on given design requirements. A
single objective constrained optimization formulation of the
toy submarine design problem is considered in this paper
and solved using two state of the art optimization algorithms
NSGA-II and IDEA. The study demonstrated the benefits
of preserving marginally infeasible solutions in IDEA that
Fig. 11: USS Dallas RC toy submarine accounts for superior performance of IDEA over NSGA-II
for constrained optimization problems. It is emphasized that
Propeller unit for Propeller unit for
Nose vertical movement lateral movement Tail
the modularity and catalogue driven structure of the proposed
framework allows for design of underwater vehicles of various
sizes, propulsions and power systems.
With the final optimum design, more and more refined anal-
ysis can be performed using external tools as a post-process.
The integration of computational fluid dynamics (CFD) tool
for better drag estimation, and an analysis module that could
Controller unit Battery compartment Propeller unit for
longitudinal movement
assist in vehicle dynamics and control analysis remain to be
performed in future works.
Fig. 12: Configuration of the USS Dallas RC toy submarine
ACKNOWLEDGEMENTS
The authors would like to thank Dr. Amitay Isaacs, Mr.
TABLE III: Performance criteria of the existing USS Dallas
Hemant K. Singh and Mr. Ahmad F. Ayob for their support
RC and optimized toy submarines
Vehicle particulars USS Dallas RC Optimized
in programming the methods.
Nose length 45 mm 49 mm
Parallel middle body length 210 mm 231 mm R EFERENCES
Tail length 95 mm 80 mm
[1] W. Wang, X. Chen, A. Marburg, J. Chase, and C. Hann, “A low-cost
Length overall 350 mm 360 mm
unmanned underwater vehicle prototype for shallow water tasks,” in Pro-
Maximum diameter 60 mm 58 mm
ceedings of the IEEE/ASME International Conference on Mechatronic
Length to diameter ratio 5.8 6.2
and Embedded Systems and Applications, MESA 2008, Beijing, China,
Maximum dimension of the inner 39.6 mm 38 mm
2008, pp. 204–209.
square
[2] K. Mohseni, “Pulsatile vortex generators for low-speed maneuvering of
Wetted surface area 0.082385 m2 0.082624 m2
small underwater vehicles,” Ocean Engineering, vol. 33, pp. 2209–2223,
Displacement volume 0.000437 m3 0.000433 m3
2006.
Mass of the displaced water 437 g 433 g
[3] M. A. Martz, “Preliminary design of an autonomous underwater vehicle
Total mass of the vehicle 430 g 428 g
using a multiple-objective genetic optimizer,” Master’s thesis, Virginia
Length of the first lever arm 45 mm 48 mm
Polytechnic Institute and State University, Blacksburg, Virginia, United
Length of the second lever arm 90 mm 90 mm
States, May 27 2008.
X-coordinate of CG -0.981462 mm -0.982488 mm
[4] J. S. Parsons, R. E. Goodson, and F. R. Goldschmied, “Shaping of
Y-coordinate of CG -0.210313 mm -0.210533 mm
axisymmetric bodies for minimum drag in incompressible flow,” J.
Z-coordinate of CG 167.083 mm 172.734 mm
Hydronautics, vol. 8, no. 3, pp. 100–107, 1974.
X-coordinate of CB 0 0
[5] M. Diez and D. Peri, “Robust optimization for ship conceptual design,”
Y-coordinate of CB 0 0
Ocean Engineering, vol. 37, pp. 966–977, 2010.
Z-coordinate of CB 163.375 mm 169.906 mm
[6] T. Lutz and S. Wagner, “Numerical shape optimization of natural
Longitudinal distance between 3.705 mm 2.828 mm
laminar flow bodies,” in Proceedings of 21st ICAS Congress, Melbourne,
CB and CG
Australia, 1998.
Nominal speed 0.5 m/s 0.5 m/s
[7] V. Bertram and A. Alvarez, “Hydrodynamic aspects of AUV design,” in
Drag (VT method) 0.0792024 N 0.0789568 N
The Fifth Conference on Computer and IT Applications in the Maritime
Drag (G&J method) 0.0800956 N 0.0798317 N
Industries (COMPIT), Oegstgeest, Netherlands, 2006, pp. 45–53.
Drag (MIT method) 0.0825858 N 0.0821771 N
[8] A. Alvarez, V. Bertram, and L. Gualdesi, “Hull hydrodynamic optimiza-
tion of autonomous underwater vehicles operating at snorkeling depth,”
Ocean Engineering, vol. 36, pp. 105–112, 2009.
performance and controllability of the designed vehicle over [9] T. Joung, K. Sammut, F. He, and S.-K. Lee, “A study on the design
optimization of an AUV by using computational fluid dynamic analysis,”
the existing one. It is also reflected that the drag of the in Proceedings of the Nineteenth (2009) International Offshore and
optimized toy submarine is lower than that of the existing Polar Engineering Conference, Osaka, Japan, 2009.
toy submarine for the same marginal speed. [10] T. Hyakudome, T. Aoki, T. Murashima, S. Tsukioka, H. Yoshida,
H. Nakajoh, T. Ida, S. Ishibashi, and R. Sasamoto, “Key technologies
for AUV URASHIMA,” in Oceans Conference Record (IEEE), vol. 1,
V. C ONCLUSIONS 2002, pp. 162–166.
Introduced in this paper is an optimization framework [11] K. Vestgard, R. Hansen, B. Jalving, and O. Pedersen, “The HUGIN 3000
survey AUV - design and field results,” in Proceedings of the Eleventh
for the preliminary design of a small-scale, low-cost, light- (2001) International Offshore and Polar Engineering Conference, Sta-
weight toy submarine for recreational purposes. The use of an vanger, Norway, vol. 4, 2001, pp. 679–684.
[12] J. Jalbert, J. Baker, J. Duchesney, P. Pietryka, W. Dalton, D. Blidberg, IEEE Transactions on, vol. 6, pp. 182–197, 2002.
S. Chappell, R. Nitzel, and K. Holappa, “A solar-powered autonomous [16] H. K. Singh, A. Isaacs, T. Ray, and W. Smith, “Infeasibility driven
underwater vehicle,” in Oceans Conference Record (IEEE), vol. 2, 2003, evolutionary algorithm (IDEA) for engineering design optimization,” in
pp. 1132–1140. 21st Australasian Joint Conference on Artificial Intelligence: Advances
[13] H. Jackson, MIT Professional Summer Course Submarine Design in Artificial Intelligence, Auckland, NewZealand (2008), Springer-Verlag
Trends, 1992. Berlin Heidelberg, no. LNAI 5360, 2008, pp. 104–115.
[14] T. Gillmer and B. Johnson, Introduction to Naval Architecture, 2nd ed. [17] T. Ray, H. K. Singh, A. Isaacs, and W. Smith, Infeasibility Driven Evo-
US Naval Institute Press, 1982. lutionary Algorithm for Constrained Optimization, E. Mezura-Montes,
[15] K. Deb, A. Pratap, S. Agarwal, and T. Meyarivan, “A fast and elitist Ed. Constraint Handling in Evolutionary Optimization, Studies in
multiobjective genetic algorithm: NSGA-II,” Evolutionary Computation, Computational Intelligence Series 198, 2009.

View publication stats

You might also like