Algorithm 2 Recognition Algorithm.
1:	INPUT
2:	$T_{est}=\{z_{l+1},z_{l+2}\cdots ,z_m\}$, $x_{i,j}$, $i=1,2,\cdots ,l$ and $j=1,2\cdots ,m_c$;
3:
4:	OUTPUT
5:	$\widehat{Z_{test}}$;
6:
7:	Let $f_i(t=0)=-0.2$, $g_j=-0.2$, where $i=1,2,\cdots ,l$ and $j=1,2\cdots ,m_c$;
8:	for each $z_k\in T_{est}$ do
9:	Let $e_{f,k}\left(t\right)=1$;
10:	Compute the stationary solution $f_i^{\ast }$, $i=1,2,\cdots ,l$ of the Equation (1);
11:	Compute the stationary solution $g_j,j=1,2\cdots ,m_c$ of the Equation (10);
12:	while $A_{esn}\ne 1$ do
13:	Adjust the scale ${\sigma }_1$ by Algorithm 1;
14:	Compute the stationary solution $g_j,j=1,2\cdots ,m_c$ of the Equation (10);
15:	end while
16:	Let $\widehat{z_k}=z^{\left(r\right)}$, where r is the category pattern that stored by the senior neuron with the highest activation;
17:	end for
18:	$\widehat{Z_{test}}=\left\{\widehat{z_k}\right\}$, $k=l+1,l+2,\cdots ,m$.