Stability Analysis for Delayed Neural Networks: Reciprocally Convex Approach

Table 1

Allowable maximum time delay ℎ₂ under different ℎ₁ and 𝛾 in this paper along with results from other papers for comparison.

Methods	ℎ1	𝛾=0.8	𝛾=0.9	Unknown 𝛾
[37]	ℎ1=0	ℎ2=2.3534	ℎ2=1.605	ℎ2=1.5103
[37]	ℎ1=1	ℎ2=3.2575	ℎ2=2.4769	ℎ2=2.3606
[37]	ℎ1=2	ℎ2=4.2552	ℎ2=3.4769	ℎ2=3.3606
[23]	ℎ1=0	ℎ2=1.2281	ℎ2=0.8636	ℎ2=0.8298
[24]	ℎ1=0	ℎ2=1.6831	ℎ2=0.1493	ℎ2=1.088
[38]	ℎ1=0	ℎ2=2.8854	ℎ2=1.9631	ℎ2=1.781
	ℎ1=0	ℎ2=5.088(Corollary 7)	ℎ2=4.881(Corollary 7)	ℎ2=5.023(Corollary 9)
	ℎ1=1	ℎ2=4.751(Theorem 2)	ℎ2=4.066(Theorem 2)	ℎ2=5.22(Corollary 5)
	ℎ1=2	ℎ2=4.689(Theorem 2)	ℎ2=4.377(Theorem 2)	ℎ2=4.788(Corollary 5)