We introduce a stochastic optimization technique to obtain self-trapped beams in the generalized nonlinear Schrödinger equation. The method is based on combining a variational approach with a bio-inspired method: the cuckoo search algorithm that relies on Lévy flights. The proposed technique can be easily adapted to generate diverse self-trapped structures in a plethora of nonlinear media. Unlike the standard variational technique and some of the numerical algorithms previously reported, this algorithm allows for the optimization of different families of self-trapped beams concurrently.

中文翻译：

---

在非线性薛定谔方程中产生自陷梁的仿生方法

我们引入了一种随机优化技术来获得广义非线性薛定谔方程中的自陷梁。该方法基于将变分方法与仿生方法相结合：依赖 Lévy 飞行的布谷鸟搜索算法。所提出的技术可以很容易地适应以在过多的非线性介质中生成不同的自陷结构。与之前报道的标准变分技术和一些数值算法不同，该算法允许同时优化不同系列的自陷光束。