A Novel Approach to Improve the Performance of Evolutionary Methods for Nonlinear Constrained Optimization

Table 4

The performance of the proposed approach in comparison with [10] on eleven test problems. The number of function evaluations is the same in both studies. The best results are indicated in boldface.

		Best	Mean	Worst	Optimal
G1	Proposed approach	−14.99145	−14.96119	−14.81634	−15.0
	[10]	−14.7207	−14.4609	−14.0566
G2	Proposed approach	0.78727	0.74244	0.67530	0.803553
	[10]	0.79506	0.79176	0.78427
G3	Proposed approach	0.98704	0.92063	0.72812	1.0
	[10]	0.9983	0.9965	0.9917
G4	Proposed approach	−30665.259	−30662.639	−30648.807	−30655.5
	[10]	−30662.5	−30643.8	−30617.0
G5	Proposed approach	—	—	—	5126.4981
	[10]	—	—	—
G6	Proposed approach	−6917.85904	−6862.02084	−6425.38018	−6961.8
	[10]	−6901.5	−6191.2	−4236.7
G7	Proposed approach	24.52525	26.12999	29.24032	24.306
	[10]	25.132	26.619	38.682
G8	Proposed approach	0.09582504	0.09582504	0.095825036	0.0958250
	[10]	0.095825	0.0871551	0.0291434
G9	Proposed approach	680.74163	681.00480	681.53181	680.63
	[10]	681.43	682.18	682.88
G10	Proposed approach	7132.98320	7543.48592	8845.85330	7049.33
	[10]	7215.8	9141.7	11894.5
G11	Proposed approach	0.75	0.75085	0.75655	0.75
	[10]	0.75	0.75	0.75