Intégration approximative

Approximate integration is needed when the exact value of a definite integral cannot be calculated. This occurs when a function's antiderivative is unknown or cannot be expressed in closed form, or when the function is defined by empirical data points from experiments rather than a continuous formula. Numerical methods then provide the only practical solution.

Riemann sums divide the integration interval into n equal subintervals of width Δx. For each subinterval, a rectangle is constructed with height determined by the function's value at a selected point. Adding the areas of all rectangles produces a numerical approximation of the definite integral, using basic geometry to estimate integrals of complex or unknown functions.

In left endpoint approximation, each rectangle's height is set by the function value at the left end of the subinterval. For increasing functions, this underestimates area; for decreasing functions, it overestimates. Right endpoint approximation uses the right end, producing opposite results: overestimation for increasing functions and underestimation for decreasing ones.

For an increasing function, the value at the left end of each subinterval is lower than values elsewhere in that subinterval. Using this smaller value as the rectangle height creates rectangles that do not fully capture the area beneath the curve, resulting in systematic underestimation of the total integral.

When a function is defined by discrete experimental data points rather than a continuous formula, traditional symbolic integration is impossible. Approximate integration techniques like Riemann sums become essential, allowing scientists and engineers to estimate definite integrals from finite measurements collected during experiments or observations.

The width Δx of each subinterval directly affects approximation accuracy. Smaller subintervals create more rectangles, allowing finer resolution of the curve's behavior and generally producing more accurate estimates. As the number of subintervals increases and Δx decreases, the Riemann sum approximation converges toward the true value of the definite integral.

Approximate integration techniques are foundational tools in numerical analysis, especially valuable when integrals must be evaluated from discrete measurements or for functions too complex for symbolic integration. They enable practical solutions in engineering and scientific contexts where exact analytical methods are infeasible or impossible.