Figure 5
The figure contains four subplots arranged in a two-by-two grid, each sharing the same horizontal axis label “Training Time (seconds)” with tick marks at 30, 150, and 1800 seconds. The top row plots “Percent Error” on the vertical axis from 0 to 1.5, while the bottom row plots “Time Elapsed (seconds)” on the vertical axis from 0 to 700 in the left subplot and 0 to 600 in the right subplot, all with multiple lines representing combinations of “Gen” (number of generations) and “Pop” (population size): “Gen: 3, Pop: 5”, “Gen: 3, Pop: 10”, “Gen: 5, Pop: 5”, “Gen: 5, Pop: 10”, “Gen: 10, Pop: 5”, and “Gen: 10, Pop: 10”. Top left plot “(a) Training Time versus Error when upper bounds for P and I gains are 250”: The plot is titled “Error versus Training Time (UpperBound equals 250)”. For “Gen: 3, Pop: 5”, the line starts near (30, 1.2), rises slightly to about (150, 1.25), and decreases to around (1800, 1.05). For “Gen: 3, Pop: 10”, the line increases gently from roughly (30, 0.75) to (150, 0.8) and (1800, 0.85), remaining the lowest error among all settings. For “Gen: 5, Pop: 5”, error rises from about (30, 1) to a peak near (150, 1.3) before dropping to around (1800, 1). For “Gen: 5, Pop: 10”, error decreases from about (30, 1) to (150, 0.82) and then climbs to just under (1800, 0.9). For “Gen: 10, Pop: 5”, the error rises from around (30, 1) to a peak near (150, 1.15) before dropping to around (1800, 1). For “Gen: 10, Pop: 10”, the line is nearly flat, hovering around (30, 0.85), dipping slightly to (150, 0.82)“, and ending around (1800, 0.85). Top right plot “(b) Training Time versus Error when upper bounds for P and I gains are 500”: The plot is titled “Error versus Training Time (UpperBound equals 500)”. For “Gen: 3, Pop: 5”, the line starts near (30, 0.45), peaks around (150, 1.15), and then falls to about (1800, 0.56). For “Gen: 3, Pop: 10”, the line rises slightly from roughly (30, 0.35) to (150, 0.38) and ends near (1800, 0.32). For “Gen: 5, Pop: 5”, error increases gently from about (30, 0.38) to (150, 0.43) and then decreases to around (1800, 0.34). For “Gen: 5, Pop: 10”, error decreases from about (30, 0.42) to (150, 0.37) and rises again to roughly (1800, 0.48). For “Gen: 10, Pop: 5”, the line falls from around (30, 0.75) to (150, 0.42) and ends near (1800, 0.36). For “Gen: 10, Pop: 10”, the line is nearly flat, moving only slightly from about (30, 0.32) to (150, 0.34) and finishing near (1800, 0.33). Bottom left plot “(c) Time Elapsed versus Time Elapsed when upper bounds for P and I gains are 250”: The plot is titled “Error versus Training Time (UpperBound equals 250)”. For “Gen: 3, Pop: 5”, the line starts near (30, 70), dips slightly to about (150, 65), and then rises to around (1800, 240). For “Gen: 3, Pop: 10”, the line increases from roughly (30, 80) to (150, 100) and ends near (1800, 470). For “Gen: 5, Pop: 5”, time elapsed grows from about (30, 110) to (150, 120) and then to approximately (1800, 230). For “Gen: 5, Pop: 10”, the line increases from around (30, 90) to (150, 130) before rising steeply to about (1800, 690). For “Gen: 10, Pop: 5”, elapsed time rises from about (30, 115) to (150, 135) and then to roughly (1800, 255). For “Gen: 10, Pop: 10”, the line starts near (30, 200), increases to about (150, 265), and finishes around (1800, 610). Bottom right plot “(d) Time Elapsed versus Time Elapsed when upper bounds for P and I gains are 500”: The plot is titled “Error versus Training Time (UpperBound equals 500)”. For “Gen: 3, Pop: 5”, the line starts near (30, 40), rises slightly to about (150, 45), and then increases to around (1800, 95). For “Gen: 3, Pop: 10”, the line increases from roughly (30, 50) to (150, 60) and ends near (1800, 205). For “Gen: 5, Pop: 5”, time elapsed grows from about (30, 60) to (150, 70) and then to approximately (1800, 135). For “Gen: 5, Pop: 10”, the line increases from around (30, 75) to (150, 90) before rising steeply to about (1800, 315). For “Gen: 10, Pop: 5”, elapsed time rises from about (30, 65) to (150, 80) and then to roughly (1800, 250). For “Gen: 10, Pop: 10”, the line starts near (30, 140), increases to about (150, 170), and finishes around (1800, 560). Note: All the numerical data values are approximated.Results of the genetic algorithm parameter optimization experiment