OPTIMIZATION
TECHNIQUES
1
DEFINITION
The term Optimize is to make perfect.
It is defined as follows: choosing the best
element from some set of available alternatives.
An art, process, or methodology of making
something (a design, system, or decision) as
perfect, as functional, as effective as possible.
2
WHY OPTIMIZATION IS
NECESSARY
Reduce the cost
Save the time
Safety and reduce the error
Reproducibility
Innovation and efficacy
3
ADVANTAGES
Yield the best solution within
the domain of study.
Require fewer experiments to
achieve an optimum formulation.
Can trace and rectify problem in
a remarkably easier manner.
4
OPTIMIZATION
PARAMETERS
CONSTRAINED
PROBLEM TYPE
UNCONSTRAIN
ED
PARAMETERS
VARIABLES DEPENDENT
INDEPENDEN
T
5
PROBLEM TYPE
Unconstrained
In unconstrained optimization problems there are no
restrictions.
For a given pharmaceutical system one might wish to make the
hardest tablet possible.
The making of the hardest tablet is the unconstrained
optimization problem.
Constrained
The constrained problem involved in it, is to make the hardest
tablet possible, but it must disintegrate in less than 15 minutes.
6
VARIABLES
Independent variables : The independent variables are
under the control of the formulator. These might
include the compression force or the die cavity filling or
the mixing time.
Dependent variables : The dependent variables are the
responses or the characteristics that are developed due
to the independent variables. The more the variables
that are present in the system the more the complications
that are involved in the optimization.
7
8
CLASSICAL OPTIMIZATION
Classical optimization is done by using the calculus to basic
problem to find the maximum and the minimum of a
function.
The curve in the fig represents the relationship between the
response Y and the single independent variable X and we can
obtain the maximum and the minimum. By using the calculus
the graphical represented can be avoided. If the relationship,
the equation
for Y as a function of X, is available [Eq]
Graphic
Y = f(X) location of optimum (maximum or minimum)
9
DRAWBACK
Applicable only to the problems that
are not too complex.
They do not involve more than two
variables.
For more than two variables
graphical representation is
impossible. 10
APPLIED OPTIMIZATION
METHODS
EVOLUTIONARY
OPERATION
SIMPLEX METHOD
SEARCH METHOD
LAGRANGIAN METHOD
CANONICAL ANALYSIS
11
EVOLUTIONARY
OPERATION
It is the method of experimental optimization
Small changes in the formulation or process are
made (i.e. repeats the experiment so many times)
& statistically analyzed whether it is improved.
It continues until no further changes takes place
i.e., it has reached optimum-the peak
The result of changes are statistically analysed.
12
This technique is especially well suited to
a production situation
Applied mostly to TABLETS.
13
Example
Tablet
How can we get hardness
By changing
the
concentration
of binder
Hardness Response
In this example, A formulator can
changes the concentration of binder and
get the desired hardness.
Advantages
Generates information on product
development.
Predict the direction of improvement.
Help formulator to decide optimum conditions
for the formulation and process.
Limitation
More repetition is required
Time consuming
Not efficient to finding true optimum
Expensive to use.
SIMPLEX METHOD
It is an experimental method applied for pharmaceutical
systems
Technique has wider appeal in analytical method other than
formulation and processing
Simplex is a geometric figure that has one more point than the
number of factors.
It is represented by triangle.
It is determined by comparing the
magnitude of the responses after
each successive calculation.
16
Types of simplex method
Two types
Simple
x
method
Basic Modified
simplex simplex
method method
Advantages
This method will find the true optimum of a
response with fewer trials than the non-systematic
approaches or the one-variable-at-a-time method.
Disadvantages
There are sets of rules for the selection of the
sequential vertices in the procedure.
Require mathematical knowledge.
OPTIMIZATION USING FACTORIAL
DESIGNS
In factorial designs, levels of factors are
independently varied, each factor at two or
more levels.
The optimization procedure is facilitated by
construction of an equation that describes the
experimental results as a function of the factor
levels.
A polynomial equation can be constructed, in
the case of a factorial design, where the
coefficients in the equation are related to the
effects and interactions of the factor
19
Types of Experimental Design1-4
Completely Randomized Design
Randomised block Design
Full Factorial
Factorial Design Design
Fractional Factorial Design
Response surface design Central Composite Design
Box- Behnken
Design
Three level full factorial design
20
Factorial Design
For evaluation of multiple factors simultaneously.
23 means 2 is level while 3 is factor
Factorial Design is divided into two types-
- Full Factorial Design
- Fractional factorial design
4/20/17 OPTIMIZATION TECHNIQUES 21
Full Factorial Design
Simplest design to create, but extremely
inefficient
Each factor tested at each condition of the
factor
Number of runs (N)
N = yx
Where, y = number of levels,
x = number of factors
E.g.- 3 factors, 2 levels each,
N = 23 = 8 runs
4/20/17 OPTIMIZATION TECHNIQUES 22
2X Design
2 = Level
X = Input Factors
Number of Number of
factors runs
2 4
x2
3 8
x3
4 16
x1
5 32
4/20/17 OPTIMIZATION TECHNIQUES 23
The equation constructed from a 2 n factorial
experiment is of the following form:
Where Y is the measured response, Xi is the level
(e.g., concentration) of the ith factor, Bi, Bij ,, Bijk ,
represent coefficients computed from the
responses of the formulations in the design
24
Suppose we already have the equation
25
26
Coeff . B1
Coeff . B2
B0 is the sum of all of the observations, Y, divided by 8 = 10.725
27
The final polynomial equation for predicting the
response,Y,is
The above equation looks entirely different from the
previous one , but both give the same response.
Above equation uses the transformed levels ofX1,
X2, and X3(+1or -1), and previous uses the actual,
observed, untransformed values.
28
Example
In this experiment, a combination drug product was
tested to obtain the dose of each drug which would
result in an optimal response.
The product contained two drugs, A(X1) and B(X2).
The experiment consists of formulating combinations
containing each drug at two dose levels.
The doses for A were 5 mg and 10 mg; B was chosen
at doses of 50 mg and 100 mg.
These levels were carefully selected to cover a range
of doses which would include an appropriate dose to
be chosen as the prime candidate for the final
marketed product.
The product is a local anesthetic, and the response
(Y) is the average time to anesthesia for 12 patients
per group
29
The high and low levels of drug A and drug B are
transformed to +1 and -1 . For drug A, the
transformation is
X the average of the two
levels
one-half the difference of the
levels
30
Coeff .
B1
Formulation 4, which has the high levels of both
drugs, has the shortest time to anesthesia, and
formulation 1 or 4 would be chosen as optimal if
either a long time or a short time to anesthesia is
desired.
31
Suppose that a time of 5 min is the most desirable
time based on considerations such as the
administration of the product and the type of
conditions that are meant to be treated with the aid
of the product
Based on a time to anesthesia of approximately 5
min, a formulation containing 0.5 of A and 1ofB
would be a candidate.
Decoding the values result in a formulation
32
containing 8.75 mg of A and 100 mg of B.
simplex lattice or mixture
designs
In order to design the best formulation, use
a trial and error approach is not an effective
way.
The simplex lattice or mixture designs
technique was applied for designing the
various formulations.
In simplex lattice, given equation is used for
design the formulations and predict the
responses:
Y = 1 (A)+ 2 (B)+ 12 (A) (B)
33
In the proposed work A is hydrophilic granules
part and B is hydrophobic granules part.
Response Y is dissolution parameters like T50%,
T80% and mean dissolution time (MDT).
1 = response at proportion of A is 100%,
2 = response at proportion of B is 100% and
12 = 4 (response at proportion of A and B is
50-50%) 2 (sum of response of A and B at
100%).
For experimental design, three formulations (i)
tablets with 100% hydrophilic granules (F-1), (ii)
tablets with 50% hydrophilic + 50% hydrophobic
granules (F-3) and (iii) tablets with 100%
hydrophobic granules (F-5) were formulated.
Then simplex lattice equation was derived for
each response using the experimental values of 34
Simplex lattice equation was derived
for each response using experimental
values of F-1, F-3 and F-5. Simplex
lattice equations for dissolution
parameters are given below:
T50% = Y1 = 3.58 (A) + 8.79 (B) - 3.3
(A) (B)
T80% = Y2 = 8.32 (A) + 20.42 (B) -
7.72 (A) (B)
MDT = Y = 4.88 (A) + 8.95 (B) +
35
Two check point formulations F-2
(A 70% and B 30%) and F-4 (A 30% and B 70%) were
formulated and characterize for all in vitro dissolution
parameters. Formulation F-2 and F-4 were formulated
and characterized for all in vitro dissolution
parameters
and responses were also predicted using simplex
lattice
equations.
36
Formulations MDT T50% T80%
A B (h) (h) (h)
F-1 100 0 4.88 3.11 10.51
F-2 70 30 5.94 4.31 11.56
F-2 (predicted) 70 30 11.33
F-3 50 50 7.02 5.33 13.73
F-4 30 70 8.31 6.77 16.03
F-4 (predicted) 30 70 7.80 6.54 15.96
F-5 0 100 8.95 9.11 20.72
37
38
