Research Article
Modelling and Output Power Estimation of a Combined Gas Plant and a Combined Cycle Plant Using an Artificial Neural Network Approach
Table 7
The training algorithms’ numerical comparison.
| S/N | SCG algorithm | LM algorithm | BR algorithm |
| 1 | After 139 iterations, the network has been trained, computationally expensive | After 44 iterations, the network has been trained, with a more beneficial computational cost | After 307 iterations, the network has been trained, most expensive computational cost | 2 | Highest MSE value of 17.62 | Permissible and lower MSE value of 16.92 | Optimum (lowest) MSE value of 16.3 | 3 | Very good regression values (R) of 0.970 | Better regression values (R) of 0.971 | More robust (higher) regression values (R) of 0.972 | 4 | Poor error histogram, due to the larger peak deviations (targets and actual data) | Satisfactory error histogram, due to lower peak deviations | Very good error histogram, due to the lowest peak deviations |
|
|