Abstract: A new constraint of clustering for the AMD calculation is proposed. This new constraint gives us large improvement in studying the cluster structure by AMD which sometimes meets difficulty in giving rise to some specific cluster configurations. The usefulness of this new constraint is verified by applying it to the the study of the third $K^\pi = 0^+$ band of $^{20}$Ne which has been discussed to have $2\alpha$-$^{12}$C structure. This band has not been easy even to construct by AMD. We see that the AMD+GCM calculation by the use of the new constraint gives rise to the third $K^\pi = 0^+$ band which contains the $2\alpha$-$^{12}$C structure as an important component.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is **how to stabilize near - nonhyperbolic chaotic systems**, and explore their potential applications in neuroscience.
Specifically, traditional Ott - Grebogi - Yorke (OGY) methods and their variants rely on linearization and local stability analysis when controlling chaotic systems. However, for near - nonhyperbolic chaotic systems, these methods fail due to the nonlinearity of the systems and the lack of prior analytical knowledge. Such systems occur naturally in fields such as astrophysics and neurobiology, but existing control methods cannot handle them effectively.
To solve this problem, the author proposes a feedback control scheme based on the invariance principle of differential equations, which has the following characteristics:
1. **No prior analytical knowledge required**: This method can stabilize chaotic systems without relying on the specific analytical expressions of the systems.
2. **Variable feedback intensity**: By introducing a variable feedback intensity \( \epsilon \), the control strategy becomes more flexible and robust.
3. **Applicable to near - nonhyperbolic systems**: Especially for those near - nonhyperbolic chaotic systems that cannot be handled by the OGY method, such as the Hindmarsh - Rose neuron model and the Rössler hyper - chaotic system.
Through this method, the author has successfully stabilized the Hindmarsh - Rose neuron model and the Rössler hyper - chaotic system, and demonstrated the effectiveness and wide applicability of this method.
### Key Formulas
- The dynamic equation of the system:
\[
\dot{x}=f(x)
\]
where \( x=(x_1, x_2, \cdots, x_n)\in\mathbb{R}^n \), \( f(x)=(f_1(x), f_2(x), \cdots, f_n(x)):\mathbb{R}^n\rightarrow\mathbb{R}^n \) is a nonlinear vector function.
- The feedback control law:
\[
\dot{x}=f(x)+\epsilon x
\]
where \( \epsilon = (\epsilon_1, \epsilon_2, \cdots, \epsilon_n) \) is the feedback intensity.
- The feedback intensity update law:
\[
\dot{\epsilon}_i = -\gamma_i x_i^2,\quad i = 1, 2, \cdots, n
\]
where \( \gamma_i>0 \) is an arbitrary constant.
- Lyapunov function:
\[
V=\frac{1}{2}\sum_{i = 1}^n x_i^2+\frac{1}{2}\sum_{i = 1}^n\frac{1}{\gamma_i}(\epsilon_i + L)^2
\]
where \( L \) is a constant greater than \( nl \), and \( l \) is a constant that satisfies the condition \( |f_i(x)|\leq l\max_j|x_j| \).
Through the above method, the author not only solves the problem of stabilizing near - nonhyperbolic chaotic systems, but also provides a new perspective for understanding the role of chaotic components in the brain, especially in explaining the multi - functional flexibility of higher life forms.