Table 5 An evaluation of the accuracy of y (t) using NDsolve and ANN-PSO-NNA for case 1, with \({\varvec{\sigma}}\) = 0.1, R = 0.2 and B = 0.3.
t | \({x(t)}_{Numerical}\) | \({\widehat{x}(t)}_{ANN}\) | \({AE(y(t))}_{ANN}\) |
|---|---|---|---|
0 | 1 | 0.999997313 | 2.69E−06 |
0.1 | 0.904930603 | 0.904960789 | 3.02E−05 |
0.2 | 0.819077383 | 0.819086889 | 9.51E−06 |
0.3 | 0.741542942 | 0.741558898 | 1.60E−05 |
0.4 | 0.671516632 | 0.67155452 | 3.79E−05 |
0.5 | 0.608266533 | 0.608300919 | 3.44E−05 |
0.6 | 0.551132369 | 0.551137058 | 4.69E−06 |
0.7 | 0.499518461 | 0.499497847 | 2.06E−05 |
0.8 | 0.452887662 | 0.452874396 | 1.33E−05 |
0.9 | 0.410755511 | 0.410778667 | 2.32E−05 |
1 | 0.372685019 | 0.372719378 | 3.44E−05 |
