End-point corrections for the midpoint rule
Provides a more accurate and stable alternative to Newton-Cotes rules for numerical integration, a fundamental problem in scientific computing.
The paper introduces a new family of numerical integration rules that achieve up to half the error of Newton-Cotes rules with better numerical stability, formulated as midpoint rule with a correction term for cheap error estimation.
In this paper we present a new family of rules for numerical integration. This family has up to half the error of the widely used Newton-Cotes rules when a sufficient number of points is evaluated and also much better numerical stability for high orders. These rules can be written as the midpoint rule with a correction term, providing a straightforward and computationally cheap way to obtain error estimations. The rules are interpolatory and use evenly spaced points, which makes them well suited for many practical applications. Their major potential disadvantage is the use of points outside the integration interval.