Introduction and Basic

Components
(iii) Optimization and Simulation

Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Objectives

 Understand the concept of optimization and simulation

 Classification of optimization and simulation problems
 Economic criteria in water resources
 Learn about the challenges in water resources

2 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Modelling Techniques
 Modelling or system analysis techniques - Developed during the Second

World War to deploy limited resources in an optimum manner

 These techniques were aided for military operations - known as operation

research techniques
 Popular operations research techniques include
 Optimization methods
 Simulation
 Game theory
 Queuing theory etc

 Among, these, the popular ones in water resources field are optimization

and simulation.

3 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Optimization
 Science of choosing the best amongst a number of possible alternatives

 Identify the best through evaluation from a number of possible solutions

 Driving force in the optimization is the objective function (or functions)

 Optimal solution is the one which gives the best (either maximum or
minimum) solution under all assumptions and constraints
 An optimization model can be stated as:

Objective function: Maximize (or Minimize) f(X)

Subject to the constraints
gj(X) ≥ 0, j = 1,2,..,m
hj(X) = 0, j = m+1, m+2,.., p
X is the vector of decision variables; g(X) are the inequality constraints;
h(X) are the equality constraints.
4 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
Classification of Optimization Techniques

Optimization problems can be classified based on the

 Type of constraints

 Nature of design variables

 Physical structure of the problem

 Nature of the equations involved

 Permissible value of the design variables

 Deterministic/ Stochastic nature of the variables

 Separability of the functions and number of objective functions

5 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
Classification based on existence of constraints

 Constrained optimization problems: Subject to one or more

constraints
 Unconstrained optimization problems: No constraints exist

6 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
Classification based on physical structure of the problem

Optimization problems are classified as optimal control and non-optimal

control problems.
(i) Optimal control problems
 Problem involving a number of stages

 Each stage evolves from the preceding stage in a prescribed manner.

 Defined by two types of variables: the control or design and state variables.

 Control variables define the system and controls how one stage evolves into the

next
 State variables describe the behavior or status of the system at any stage

 Problem is to find a set of control variables such that the total objective function

(also known as the performance index, PI) over all stages is minimized, subject
to a set of constraints on the control and state variables.
7 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
Classification based on the physical structure of the problem…
An OC problem can be stated as follows:
Find X which minimizes
Subject to the constraints

where xi is the ith control variable, yi is the ith state variable, and fi is the
contribution of the ith stage to the total objective function. gj, hk, and qi are the
functions of xj, yj ; xk, yk and xi, yi , respectively, and l is the total number of
states.
(ii) Problems which are not optimal control problems are called non-optimal
control problems.
8 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
Classification based on the nature of the equations
involved
Optimization problems can be classified as

(i) Linear programming: Objective function and all the constraints are
‘linear’ functions of the design variables

(ii) Nonlinear programming : Any of the functions among the objectives

and constraint functions is nonlinear

(iii) Geometric programming : Objective function and constraints are

expressed as polynomials

(iv) Quadratic programming: Best behaved nonlinear programming problem

with a quadratic objective function and linear constraints and is concave

(for maximization problems)
* For details refer lecture notes
9 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Classification based on the permissible values of the
decision variables

Objective functions can be classified as integer and real-valued

programming
(i) Integer programming problem: Some or all of the design variables of an
optimization problem are restricted to take only integer (or discrete)
values

(ii) Real-valued programming problem: Minimize (or maximize) a real

function by systematically choosing the values of real variables from
within an allowed set. When the allowed set contains only real values, it is
called a real-valued programming problem
10 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Classification based on deterministic/ Stochastic nature
of the variables

Optimization problems can be classified as deterministic or stochastic

programming problems

(i) Deterministic programming problem: In a deterministic system, for the

same input, the system will produce the same output always. In this type of
problems all the design variables are deterministic.

(ii) Stochastic programming problem: In this type of problem, some or all

the design variables are expressed probabilistically (non-deterministic or
stochastic).

11 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Classification based on separability of the functions
Optimization problems can be classified as separable and non-separable
programming problems

(i) Separable programming problems: In this type of problem, the objective

function and the constraints are separable.

Function is said to be separable if it can be expressed as the sum of n

single-variable functions,
f 1 xi , f 2 x 2 ,... f n x n  ,
n
i.e.f ( X )  f i xi 
i 1

(ii) Non-separable programming problems: Objective function is not

separables

12 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Classification based on the number of objective
functions
Objective functions can be classified as single-objective and multi-objective
programming problems.

(i) Single-objective programming: There is only a single objective function.

(ii) Multi-objective programming: A multiobjective programming problem

can be stated as follows:

Find X which maximizes/ minimizes f1 X , f 2 X ,... f k X 

Subject to gj(X) ≤ 0 , j = 1, 2, . . . , m

where f1, f2, . . . fk denote the objective functions to be maximized/

minimized simultaneously

13 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Simulation
 Simulation process duplicates the system’s behaviour by designing a model of the

system and conducting experiments for a better understanding of the system

functioning in various probable scenarios
 Simulation reproduces the response of the system to any imposed future conditions

 Main advantage of simulation is its ability to accurately describe the reality

 Operating policies can be tested through simulation before implementing in actual

situations
 Water resources systems are too complex to be expressed in any analytical

expression.
 Simulation model duplicates the system’s operation with a defined operational

policy, parameters, time series of flows, demands etc

 Design parameters and the operation policy are evaluated through the objective

function or some reliability measures.

14 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Steps in Simulation
1. Problem definition: Define the goals of the study

2. System definition: Identify the water resources system components and its

hydrological aspects. Identify the performance measures to be analysed.

3. Model design: Understand the behavior of actual system. Decide the model

structure by determining the variables describing the system, its interaction

and various parameters of structures. Decide the inputs (time series of
flows, demands of the system, operation policies etc) and outputs
(hydrological variables and design variables).
4. Data Collection: Determine the type of data to be collected. New/ Old data

is collected/ gathered.
5. Validation: Test the model and apply the model to the problem
15 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Classification of Simulation models
 Simulation models can be

i. Physical (e.g. a scale model of a spillway)

ii. Analog (system of electrical components such as resistors or capacitors arranged

to act as an analog to the hydrological components) or

iii. Mathematical (action of a system expressed as equations or logical statements.

 Simulation models can be

i. Static (fixed parameters and operational policy) or

ii. Dynamic (takes into account the change in the parameters of the system and the

operational policy with time) in nature.

16 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Classification of Simulation models…

 Simulation models can be deterministic or stochastic

 Simulation models can be statistical or process oriented, or a mixture of both.

 Pure statistical models are based solely on data (field measurements). Regressions

and artificial neural networks are examples

 Pure process oriented models are based on knowledge of the fundamental processes

that are taking place. In this, calibration using field data is required to estimate the
parameter values in the process relationships.
 Hybrid models incorporate some process relationships into regression models or

neural networks.

17 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc
 Comparison between Optimization and Simulation
 Optimization models eliminate the worst solutions.

 Simulation tools evaluate the performance for various configurations of the system;

but they are not effective for choosing the best configuration.
 Simulation simply addresses ‘what-if’ scenarios – what may happen if a particular

scenario is assumed or if a particular decision is made. Users have to specify the

value of design or decision variables for conducting simulation.
 Simulation is not feasible when there are too many alternatives for decision

variables, which demand an enormous computational effort.

 Optimization will determine the best decision; but the solution is often based on

many limiting assumptions.

 Full advantage of systems techniques: Optimization should be used to define a

relatively small number of good alternatives that can later be tested, evaluated and
improved by means of simulation.
18 Water Resources Systems Planning and Management: M1L3 D. Nagesh Kumar, IISc