Flame Failures and Recovery in Industrial Furnaces: A Neural Network Steady-State Model for the Firing Rate Setpoint Rearrangement

Table 6

The performances of the developed neural network estimator.

	NN model	Burner 1	Burner 2	Burner 3	Burner 4	Burner 5
	Train	0.9997	0.9991	0.9969	0.9999	0.9991
	Test	0.9931	0.9992	0.9901	0.9944	0.99601
	Validation	0.9941	0.9943	0.9975	0.9987	0.9921
MSE	Train (×10^−5)	7.3233	3.2427	3.99622	36.274	73.379
	Test (×10^−4)	6.3474	17.91	22.217	6.1263	11.521
	Validation (×10^−4)	0.11646	24.157	17.34	51.057	34.751