In this paper, a novel nature-inspired optimization algorithm, hazelnut tree search (HST) is proposed for solving numerical and engineering optimization problems. HST is a multi-agent algorithm that simulates the search process for finding the best hazelnut tree in a forest. The algorithm is composed of three main actuators: growth, fruit scattering, and root spreading. In the growth phase, trees compete with each other on shared resources to grow up and improve their fitness. In the fruit scattering phase, HTS performs exploration by simulating the movement of hazelnuts around the forest with the help of animals and rodents. In the root spreading, HTS performs exploitation by modeling the root spreading mechanism of trees around themselves. The performance of the proposed algorithm is evaluated on multi-variable unconstraint numerical optimization benchmarks and constraint engineering problems. Comparing the proposed algorithm with a few other optimization algorithms shows the superiority of the HTS in terms of problem-solving success and finding the global optimum on most benchmark problems.

在本文中，提出了一种新颖的自然启发优化算法榛子树搜索（HST）来解决数值和工程优化问题。HST 是一种多智能体算法，可模拟在森林中寻找最佳榛子树的搜索过程。该算法由三个主要执行器组成：生长、果实散射和根传播。在生长阶段，树木在共享资源上相互竞争以成长并提高它们的健康度。在散果阶段，HTS 在动物和啮齿动物的帮助下，通过模拟榛子在森林中的运动来进行探索。在根传播中，HTS 通过对自身周围树木的根传播机制进行建模来进行开发。在多变量无约束数值优化基准和约束工程问题上评估了所提出算法的性能。将所提出的算法与其他一些优化算法进行比较，可以看出 HTS 在解决问题的成功和在大多数基准问题上找到全局最优方面的优越性。