ADAM’S-MOULTON METHOD

Taylor methods and Runge–Kutta (RK) methods are known as single-step or one-step
methods, since at a typical step yn+1 is determined solely from yn.
The Adams–Bashforth (AB) and Adams–Moulton (AM) methods are the basis of
some of the most widely used computer codes for solving the initial value problem. They are
generally more efficient than the RK methods, especially if one wishes to find the solution with
a high degree of accuracy or if the derivative function f(t, y) is expensive to evaluate.

To understand the Adams-Moulton method thoroughly, it's essential to be familiar with

several key terms and concepts associated with numerical methods for solving ordinary
differential equations (ODEs). Here are some important terms to know:

Multistep Methods: Numerical methods for solving ODEs that involve using information from
multiple previous time steps to compute the solution at the next time step.

Implicit Methods: Numerical methods where the solution at the next time step depends on both
the current and future time steps, requiring the solution of an implicit equation.

Adams-Bashforth Method: A family of explicit multistep methods for solving ODEs, where the
solution at the next time step is based on extrapolation from previous time steps.

Adams-Moulton Method: A family of implicit multistep methods for solving ODEs, where the
solution at the next time step is based on interpolation between the current and future time steps.

Truncation Error: The error introduced by approximating an exact mathematical procedure with
a numerical one, which decreases as the step size decreases.

Stability: The property of a numerical method that determines whether small errors in the initial
conditions or round-off errors grow or decay over time.

Convergence: The property of a numerical method that ensures that the approximate solution
approaches the true solution as the step size tends to zero.

Order of Accuracy: The rate at which the error in the numerical solution decreases as the step
size decreases, typically expressed as a power of the step size.
Step Size: The size of the time interval between consecutive time steps in the numerical
integration process.

Initial Value Problem (IVP): A differential equation along with initial conditions that specify the
value of the dependent variable at an initial time.

Boundary Value Problem (BVP): A differential equation along with boundary conditions that
specify the value of the dependent variable at boundary points.

Function Evaluation: The process of evaluating the right-hand side of the differential equation at
specific time and state values.

Here's what the term "order" means in the context of the Adams-Moulton method:

First Order: A first-order Adams-Moulton method is accurate to the first order in the step size

ℎ h. This means that the error in the numerical solution decreases proportionally to ℎ h as the
step size decreases. The basic Adams-Moulton method is typically first order.

Second Order: A second-order Adams-Moulton method is accurate to the second order in the
step size ℎh. This means that the error in the numerical solution decreases proportionally to ℎ2h
2mas the step size decreases. Higher-order Adams-Moulton methods can be derived to achieve
second-order accuracy.

Higher Orders: Similarly, higher-order Adams-Moulton methods can be developed to achieve

even greater accuracy, such as third-order, fourth-order, and so on. These methods require more
function evaluations and typically involve solving more complex implicit equations at each step.

In summary, the order of the Adams-Moulton method indicates the rate at which the error
in the numerical solution decreases as the step size decreases. Higher-order methods provide
greater accuracy but may require more computational effort to implement. The choice of order
depends on the desired level of accuracy and computational efficiency for a particular
application.

To evaluate the integral:

t n+1

∫ g ( t ) dt , g ( t ) =Y ' ( t )=f ( t , Y (t ) ) , (6.2)

tn
 we approximate g(t) by using polynomial interpolation and then integrate the
interpolating polynomial. For a given nonnegative integer q, the AB methods use interpolation
polynomial of degree q at the points {t n, tn−1, . . ., tn−q}, and AM methods, use interpolation
polynomial of degree q at the points {tn+1, tn, tn−1, . . ., tn−q+1}.

As with the AB methods, we begin our presentation of AM methods by considering the

method based on linear interpolation. Let p1(t) be the linear polynomial that interpolates g(t) at t n
and tn+1,

Using this equation to approximate the integrand in (6.2), we obtain the trapezoidal rule
discussed in Chapter 4,

Dropping the last term, the truncation error, we obtain the AM method;

This is the trapezoidal method discussed in Section 4.2. It is a second-order method and
has a global error of size O(h2). Moreover, it is stable.
Example 6.4. Solve the earlier problem (6.9) by using the AM method (6.17) (the trapezoidal
method). The results are given in Table 6.4 for h = 0.05, 2h = 0.1, and the Richardson error
estimate for second-order methods is given in the last column. In this case, the O(h 2) error
behavior is very apparent, and the error estimation is very accurate.

Example 6.5. Repeat Example 6.4, but using the procedure described following (4.28) in
Chapter 4, with only one iterate being computed for each n. Then, the errors do not change
significantly from those given in Table 6.4. For example, with t = 10 and h = 0.05, the error is;

This is not very different from the value of −1.75 × 10−4 given in Table 6.4. The use of
the iterate y(1)n+1 as the root yn+1 will not affect significantly the accuracy of the solution for most
differential equations. Stiff differential equations are a major exception.

As in Table 6.5, the column heading “T. Error” denotes “Truncation Error”. Note
that the AM method of order 1 is the backward Euler method, and the AM method of
order 2 is the trapezoidal method.

AB vs. AM

The Adams-Bashforth method and the Adams-Moulton method are both numerical
methods used for solving ordinary differential equations (ODEs). They are both members of the
family of multi-step methods for numerical integration.

Order: Adams-Bashforth method: It is an explicit method, meaning that the formula for the next
step only involves information from previous steps. The Adams-Bashforth method is typically
one order higher than the number of steps used. For example, the Adams-Bashforth 3-step
method (AB3) is third order accurate.

Adams-Moulton method: It is an implicit method, meaning that the formula for the next step
involves an implicit equation that must be solved iteratively. The Adams-Moulton method is
typically one order higher than the Adams-Bashforth method of the same number of steps. For
example, the Adams-Moulton 3-step method (AM3) is fourth order accurate.
Accuracy: Adams-Bashforth methods tend to have slightly less accuracy compared to Adams-
Moulton methods for the same number of steps, but they are easier to implement due to their
explicit nature.

Adams-Moulton methods are generally more accurate, especially for stiff ODEs, but they require
solving a nonlinear equation at each step, which can be computationally expensive.

Stability: Adams-Bashforth methods are typically more prone to stability issues, especially for
stiff ODEs, due to their explicit nature.

Adams-Moulton methods, being implicit, are generally more stable, especially for stiff problems.

Derivation:

Both methods are derived from the same underlying principles of approximating the integral in
the ODE, typically using a polynomial approximation. The main difference lies in how the
coefficients of the approximation are determined and the resulting formulas for advancing the
solution.

In summary, while both methods belong to the family of multi-step methods and share
similarities in their approach, the key differences lie in their order of accuracy, explicitness vs.
implicitness, stability characteristics, and computational cost. The choice between Adams-
Bashforth and Adams-Moulton methods depends on the specific requirements of the problem
being solved, such as accuracy, stability, and computational efficiency.

Like many numerical integration methods, the Adams-Moulton method is derived by

approximating the integral in the differential equation using a polynomial interpolation. In this
case, it is often derived from the Newton-Cotes formulas.
Examples
Why AM Method

The Adams-Moulton method, like other numerical methods, is used to approximate the
solution of a differential equation when the true solution cannot be obtained analytically or when
it is computationally expensive to do so.

Analytical Intractability: Many differential equations do not have closed-form analytical

solutions, especially for complex or nonlinear systems. In such cases, numerical methods like the
Adams-Moulton method provide a practical way to approximate the solution.

Computational Efficiency: Even for differential equations with known analytical solutions, it
may be computationally expensive or impractical to evaluate the true solution, especially for
large-scale or time-dependent problems. Numerical methods offer a more efficient way to obtain
approximate solutions within a reasonable amount of computational resources.

Flexibility: Numerical methods like the Adams-Moulton method can handle a wide range of
differential equations, including nonlinear, stiff, and higher-order systems. They provide a
flexible framework for solving differential equations in various scientific and engineering
applications.
Boundary Value Problems: The Adams-Moulton method, along with other numerical methods,
can be extended to solve boundary value problems (BVPs) and initial-boundary value problems
(IBVPs) where analytical solutions may not be available or practical.

Accuracy Control: Numerical methods allow for fine-tuning of the solution accuracy by
adjusting parameters such as the step size and the order of the method. This enables users to
obtain solutions with the desired level of accuracy while balancing computational cost.

Implementation in Software: Numerical methods like the Adams-Moulton method are

implemented in various software packages and libraries, making them readily accessible to
researchers, engineers, and practitioners. These tools provide efficient algorithms and data
structures for solving differential equations in a user-friendly manner.

Thus, the Adams-Moulton method, along with other numerical methods, serves as a
valuable tool for approximating the solution of differential equations when analytical solutions
are unavailable, computationally expensive, or impractical to obtain. Its use enables researchers
and practitioners to tackle a wide range of scientific and engineering problems efficiently and
effectively.

Application of AM Method

The Adams-Moulton method primarily provides a computationally efficient way to

approximate the solutions of ordinary differential equations (ODE), allowing researchers and
practitioners to gain insights into complex systems and make informed decisions.

Physics Simulations: Many physical systems are described by sets of ordinary differential
equations. The Adams-Moulton method can be used to numerically simulate the behavior of
these systems over time. Examples include simulating the motion of celestial bodies, such as
planets and satellites, in gravitational fields.

Engineering Analysis: Engineers often encounter ODEs when analyzing dynamic systems such
as electrical circuits, mechanical systems, and chemical reactors. The Adams-Moulton method
can be used to solve these differential equations numerically, allowing engineers to predict
system behavior and optimize designs.
Biology and Medicine: Mathematical models of biological systems, such as population
dynamics, enzyme kinetics, and pharmacokinetics, are often formulated as systems of ordinary
differential equations. The Adams-Moulton method can be applied to these models to simulate
biological processes and predict outcomes.

Climate Modeling: Climate models simulate the behavior of the Earth's atmosphere, oceans, and
land surface over time. These models involve complex systems of differential equations that
describe processes such as fluid flow, heat transfer, and chemical reactions. The Adams-Moulton
method can be used to numerically integrate these equations and simulate the Earth's climate
under different scenarios.

Financial Modeling: Differential equations are also used in finance to model the behavior of
financial instruments, such as options and derivatives, as well as economic processes such as
interest rates and inflation. The Adams-Moulton method can be applied to these models to
analyze financial markets and make predictions about future trends.

Sample Python code implementing the Adams-Moulton method for solving a first-
order ordinary differential equation (ODE) initial value problem:
https://pubs.sciepub.com/tjant/9/1/1/index.html