Вы находитесь на странице: 1из 2

Numerical Methods

Numerical methods are a mathematical tool designed to solve numerical problems. These methods are
basically algorithms used for computing numeric data. They are used to provide approximate results for
the problems being dealt with and their necessity is felt when it becomes impossible or extremely difficult
to solve a given problem analytically. The goal in numerical analysis is to develop a finite sequence of
steps, i.e., an algorithm, for computing an approximation to the value f(x). The implementation of
a numerical method with an appropriate convergence check in a programming language is called
a numerical algorithm.

Numerical methods in Chemistry deal with a broad range of problems particularly those involved in data
treatment like calculations on atomic or molecular level, quantum mechanical calculations of atoms and
molecules, numerical treatment of chemical reaction kinetics etc which involves non linear equations.
There are two methods generally used to solve non linear equations:

A. Direct Method- This method attempt to solve the problem by a finite sequence of operations. In
the absence of rounding errors, direct methods would deliver an exact solution (like solving a
linear system of equations. It gives exact values of roots at the same time in a finite number of
steps. Example- Quadratic equation, ax2+bx+c = 0, the roots are given by:

B. Approximate or Iterative Method- This is based on the idea of successive approximation i.e.,
starting with one or more initial approximation to the value of root. For any
iterative numerical technique, each successive iteration results in a solution that moves
progressively closer to the true solution. This is known as convergence. A numerical method is
not always guaranteed to produce converging results as it is subjected to satisfying certain
conditions. This method generally gives only one root at a time.

Types of Approximate Methods:

1. Iterative Method

Iterative methods are often the only choice for nonlinear equations. Iteration means repeatedly carrying
out a process. To solve an equation using iteration, start with an initial value and substitute this into
the iteration formula to obtain a new value, and then use the new value for the next substitution, and so
on. This repeated trial and error procedure successively refines an initial crude guess value till successive
values becomes too close (<0.0001) i.e., method has converged.

Example: Find the real root of the equation x3 + x 2 = 1 by iteration method.


Solution: We can rewrite the above equation by
x3 + x 2 - 1 = 0;
Let f(x) = x3 + x 2 - 1
f(0) = -1 (positive)
f(1) = 1 (negative)
Hence the root value lie between 0 to 1
x3 + x 2 - 1 = 0
x2 (x + 1) = 1
x2 = 1/ (x + 1)
x = 1/ √(x + 1)
pi(x) = 1/ √(x + 1)

Let the initial approximation be x0 = 0.5

x1 =pi(x0) = 1/√1+.5 = 0.81649


x2 =pi(x1) = 1/√1+ 0.81649 = 0.74196
x3 = pi(x2) = 1/√1+ 0.74196 = 0.75767
x4 = pi(x3) = 1/√1+ 0.75767 = 0.75427
x5 = pi(x4) = 1/√1+ 0.75427 = 0.75500
x6 = pi(x5) = 1/√1+ 0.75500 = 0.75485
x7 = pi(x6) = 1/√1+ 0.75485 = 0.75488
Since the difference between x6 and x7 are very small (<0.0001), so the final root is 0.75488.

2. Newton-Raphson method is one of the most widely used methods for root finding. It can be easily
generalized to the problem of finding solutions of a system of non-linear equations, which is referred to as
Newton's technique. Moreover, it can be shown that the technique is quadratically convergent as we
approach the root. This method requires only one initial value x, which we will refer to as the initial
guess for the root. It uses an iterative process to approach one root of a function using recursive formula
as given in equation 1. The specific root that the process locates depends on the initial, arbitrarily chosen
x-value.

(1)

Here, xn is the current known x-value, f(xn) represents the value of the function at xn, and f'(xn) is the
derivative (slope) at xn. xn+1 represent the next x-value that you are trying to find. Essentially, f'(x), the
derivative represents f(x)/dx (dx = delta-x). Therefore, the term f(x)/f'(x) represents a value of dx.

(2)

The more iteration that are run, the closer dx will be to zero. Theoretically, we could execute an infinite
number of iterations to find a perfect representation for the root of our function. Therefore iteration will
be carried out and xn+1 is calculated from recursive formula as given in equation 1 and compared with xn
and if the two values are more than 0.0001then xn+1 is substituted in r.h.s of equation 1 and calculated next
value xn+1 and so on computation is repeated till the process has worked accurately i.e., delta-x becomes
less than 0.0001.

Вам также может понравиться