13.3. Runge–Kutta Fourth-Order Method (RK4)

The RK4 method is widely used due to its excellent accuracy and relatively low computational cost. It evaluates the slope at four points in each interval and takes a weighted average.

The Runge-Kutta Fourth-Order Method (RK4) is a numerical technique used to solve ordinary differential equations (ODEs). This method is preferred because it provides a higher level of accuracy compared to simpler methods. What sets RK4 apart is that it calculates slopes at four different points within each interval of the solution. These slopes are combined with different weights to produce a more accurate estimate of the function's value at the next step.

Imagine you are trying to estimate the height of a tree from a distance. Instead of taking a single measurement, you decide to measure at multiple points: from the base, from the middle, from a point up the trunk, and finally the top. By combining these measurements, you get a more accurate estimate of the tree's height. Similarly, RK4 measures the slope of the ODE at four different locations for a more precise solution.

Algorithm Given: • Differential equation: = 𝑓(𝑥,𝑦) • Initial condition: 𝑦(𝑥 ) = 𝑦 0 0 • Step size: ℎ Steps:

1. 𝑘 = ℎ𝑓(𝑥 ,𝑦 ) 1 𝑛 𝑛 ℎ 𝑘
2. 𝑘 = ℎ𝑓(𝑥 + ,𝑦 + 1) 2 𝑛 2 𝑛 2 ℎ 𝑘
3. 𝑘 = ℎ𝑓(𝑥 + ,𝑦 + 2) 3 𝑛 𝑛 2 2
4. 𝑘 = ℎ𝑓(𝑥 +ℎ,𝑦 +𝑘 ) 4 𝑛 𝑛 3 Update solution: 1 𝑦 = 𝑦 + (𝑘 + 2𝑘 + 2𝑘 + 𝑘 ) 𝑛+1 𝑛 6 1 2 3 4

The RK4 algorithm involves several steps. First, we define the differential equation and initial conditions along with the step size (ℎ). In the algorithm, we compute four intermediate slopes (k1, k2, k3, k4) based on the function's value at various points. First, k1 is calculated at the current point. Then, k2 and k3 are computed at intermediary points, determined by k1. Finally, k4 includes the value at the end of the step. The updated solution is found by combining these slopes with weighted averages, where k2 and k3 are given more importance. This systematic approach results in a highly accurate approximation at each step.

Think of calculating the trajectory of a ball thrown into the air. Instead of just calculating its position based on its current speed and direction (like simpler methods), you also consider its potential future positions as it moves upward and downward. By evaluating how its speed changes at multiple points during its flight, you can predict its position much more accurately. This is akin to what the RK4 method does by evaluating multiple slopes.

Example (RK4): Solve = 𝑥 +𝑦, with 𝑦(0) = 1, step size ℎ = 0.1, find 𝑦(0.1). Given: 𝑓(𝑥,𝑦) = 𝑥+ 𝑦 Step 1: • 𝑘 = 0.1⋅𝑓(0,1) = 0.1(1)= 0.1 1 • 𝑘 = 0.1⋅ 𝑓(0.05,1.05) = 0.1(1.10)= 0.11 2 • 𝑘 = 0.1⋅ 𝑓(0.05,1.055)= 0.1(1.105)= 0.1105 3 • 𝑘 = 0.1⋅ 𝑓(0.1,1.1105)= 0.1(1.2105)= 0.12105 4 Final: $y_1 = 1 + \frac{1}{6}(0.1 + 2 \cdot 0.11 + 2 \cdot 0.1105 + 0.12105) \ = 1 + \frac{1}{6}(0.66205) \approx 1.1103$

In this example, we apply the RK4 method to the differential equation 𝑑𝑦/𝑑𝑥 = 𝑥 + 𝑦 with an initial condition of 𝑦(0) = 1 and a step size (ℎ) of 0.1. We compute the intermediate slopes k1 through k4 based on evaluations of the function at specific points. After calculating these values, we combine them using the formula for updating the solution, yielding an approximate value of y at 𝑥 = 0.1. This demonstrates how RK4 effectively provides a closely accurate result through multiple evaluations.

Imagine you are budgeting monthly expenses. Instead of just throwing in estimates for the month, you look through your actual spending from previous months, adjusting projections based on known costs. For instance, you might look at your rent (the first value), your grocery costs in the middle of the month, and your utility bills (the later values). By aggregating all this information thoughtfully, you can make a more realistic budget prediction — that’s similar to how RK4 pulls together different slope evaluations for accuracy.

--