Scribd
Upload
332 views4 pages

Final Exam Paper Design Optimization of Systems

This document contains details of a final examination for a course on design and optimization of energy systems. It includes 9 questions covering topics like dynamic viscosity as a function of temperature, logistic regression, solving systems of equations using Gauss-Seidel method, constrained optimization using Lagrange multipliers, Kuhn-Tucker conditions, optimization of a grain drying system using genetic algorithm and simulated annealing, linear programming to optimize aircraft engine performance, and optimal time allocation to minimize fuel consumption on a hilly road using dynamic programming. The exam is 3 hours long and carries a maximum of 100 marks.

Uploaded by

Uday Joshi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
332 views4 pages

Final Exam Paper Design Optimization of Systems

This document contains details of a final examination for a course on design and optimization of energy systems. It includes 9 questions covering topics like dynamic viscosity as a function of temperature, logistic regression, solving systems of equations using Gauss-Seidel method, constrained optimization using Lagrange multipliers, Kuhn-Tucker conditions, optimization of a grain drying system using genetic algorithm and simulated annealing, linear programming to optimize aircraft engine performance, and optimal time allocation to minimize fuel consumption on a hilly road using dynamic programming. The exam is 3 hours long and carries a maximum of 100 marks.

Uploaded by

Uday Joshi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

National Program on Technology Enhanced Learning

IIT MADRAS
Design and Optimization of Energy Systems
Final Examination
Duration: 3hr
Max. Marks: 100
1. Make suitable assumptions wherever required with justification
2. Assume any missing data
3. Graph sheet can be used for the Linear Programming problem

(1) The dynamic viscosity of liquid water in Ns/m2 varies with temperature (T in K)
as given below.
S. No
1
2
3
4

Temperature,
K
275
290
300
315

Viscosity,
Ns/m2
16.52 x 10-4
10.8 x 10-4
8.55 x 10-4
6.31 x 10-4

(a) Using Newtons divided difference method, appropriate for 4 data points,
obtain en exact fit to the viscosity as a function of temperature. (6)
(b) Using the fit obtained in (a) estimate the viscosity at 295 K
(1)
(c) Compare the result obtained in (b) with a linear interpolation.
(1)
(2) An experiment was designed to examine the effect of load, x (in appropriate
units) on the probability of failure of specimens of a certain industrial
component. The following results were obtained from the experiment.
Load, x
10
30
60
85
95
110

Number of specimens
400
500
450
600
550
350

Number of failures
17
49
112
275
310
253

The regression model suitable for this problem is of the following form (also
known as the logistic regression model)

p=

1
1+ e

( a +bx )

Where p is the probability of failure of the component. Using the above


model with the data given in the Table given above, get the best estimates of a
and b.
(10)
(b) Estimate the standard error of the estimate of p from the regression
equation and the correlation coefficient.
(2)
(3) The mass balance of three species (x1, x2 and x3 all in kg/s) in a series of
interconnected chemical reactors is given by the following equations:
1.4 x1+ 2.2 x2+ 4.8 x3 = 4.8
6 x1- 3 x2- 2x3 = 22
2 x1+ 4 x2+ x3 = 24
Using the Gauss-Seidel method with initial guess values of 1.0 for all the three
variables, determine the values of x1, x2 and x3. Perform at least 7 iterations
and report the sum of the squares of the residues of the three variables at the end
of every iteration.
(10)
(4) (a) A closed rectangular metal box is to be constructed in order to maximize the
volume. The total sheet metal available 24 m2. Using the method of Lagrange
multipliers for a constrained optimization in three variables (x1, x2 and x3- the
three dimensions of the box), determine the optimum dimensions of the box that
maximize the volume.
(10)
(b)For the above problem, report the optimal volume and the Lagrange
multiplier. Establish using second order necessary and sufficient conditions that
the optimum is indeed a maximum.
(1+1+2)
(5) (a)Write down the Kuhn-Tucker conditions for the following Non-Linear
optimization problem (NLP).
Minimize:
Subject to :

y = x12+2x22
2x1-2x2=1
x12+x22 4

(4)

(b) Solve the above problem and check whether the inequality constraint is
binding.
(4)

(6) Newly harvested grain often has a high moisture content. This must be removed
by drying to prevent spoilage and increase shelf life. The drying is usually
achieved by warming ambient air and blowing it through a bed of the grain. The
seasonal operating cost (in rupees) per square meter of grain bed for such a
dryer consists of the heating cost plus the blower cost which are given by
Heating cost: Rs. 0.09 Q T
Blower cost: Rs. 1.1 x 10-7 Q3
Where Q is the air flow rate though the bed during the season in m3/ m2 of bed
area and T is the rise in temperature through the heater in C.
The values of Q and T also influence the time required for adequate drying of
the grain according to the following equation
Drying time = 70 x 106 / Q2T days.
(a)Convert this into a single variable unconstrained optimization problem in air
flow rate, Q for a drying time of exactly 50 days.
(2)
(b) We would like to use GA to solve the above problem, as a single variable
minimization problem for the total cost in terms of the air flow rate Q. Perform
two iterations of the GA with an initial population size of 4. You may assume
that 50 Q 1023 m3/m2. An accuracy of 1 m3/m2 is sufficient to represent Q.
Decide appropriate strategies for crossover. No mutation is required. Report
maximum, minimum and average fitness values for the initial population
(which may be randomly chosen) and the populations at the end of the first and
second iterations. There is no need to convert the minimization problem to a
maximization problem. You may decide on the mating pool by giving more
preference to those candidates which yield lower total coast.
(14)
(7)Consider the problem of minimization of convective heat loss from a
cylindrical storage heater that makes use of the solar energy collected by a
suitable system. The volume of the tank is 5 m3 and is fixed. The radius of the
tank is r and the height is h.
(a) Convert this to a single variable unconstrained optimization problem in radius
r.
(1)
(b) We would like to use Simulated Annealing (SA) to solve the above problem, as
a single variable minimization problem in r. Perform four iterations of the
SA with a starting value of r = 2 m. You may assume that 0.1 r 4 m. Use
the random number table provided to you. The initial temperature T (used in

the algorithm that does not correspond to the physical temperature in this
problem) may be taken to be the average of four objective function values for
appropriately chosen values of the radius.
(9)
(8) A supersonic military aircraft engine typically employs an after burner after
the combustor where additional fuel is injected in order to increase the thrust.
The total energy, E, at the exit of the afterburner is to be maximized for
maximum work output. One such aircraft is flying at 600 m/s and intakes air at
10.5 kg/s (Cp for air is 1005 J/kgK). The temperature at the inlet of the engine
is -25C and the calorific value of the fuel is 44000 kJ/kg. The fuel flow rate to
the combustor is x1 kg/s and that to the after burner is x2 kg/s. Due to spraying
limitations, a constraint in the form of 8 x1+ 6 x2 3 arises. The combustor can
withstand more heat than the afterburner and so x1-x2 0.1. Furthermore, due
to limitations in fuel storage and distribution 2 x1 + 5 x2 1.2.
(a) Set up the optimization problem for maximizing energy at the exit (Note:
Due to high velocities, kinetic energy term may not be negligible!) of the
afterburner as a Linear Programming (LP) problem.
(4)
(b) Solve the LP using the graphical method.
(6)
(c) Solve the LP using the method of slack variables and compare the solutions
obtained in (b) and (c).
(4)
(9) Consider a container truck that is climbing a ghat (hill) road. There are three
sections on the ghat road denoted by A-B, B-C and C-D. The fuel consumed in
each section varies with the time taken to cover the particular section and is
tabulated below. Please note that section BC is the toughest part of the ghat
road that drinks so much of fuel. The hill climb needs to be completed within
a total time of 34 s. Using dynamic programming, determine the optimal time
to be spent by the truck in the three sections so as to minimize the total fuel
consumption. What is the computational gain achieved by using the dynamic
programming for this problem?
Section
A-B

B-C

C-D

Time, t, s
10
11
12
13
10
11
12
13
10
11
12
13

Fuel consumption, g
60
51
43.5
37.5
91.5
78
67.5
57
73.5
61.5
52.5
45

(8)

You might also like