ABSTRACT This paper presents an evolutionary method to find the optimally designed gripper configuration for the automated material handling process. The optimal design of the robot gripper is a non-linear, multicriteria and multi-constraint problem. Evolutionary computational methods are introduced to overcome the difficulty associated while finding the optimal dimensions of the gripper. In this study, an Improved Multi-objective Antlion Optimisation Algorithm (I-MOALO) has been proposed to find the optimal dimensions of three different robotic gripper mechanisms. The multi-objective optimisation problem (MOOP) is formulated from a geometric model of the gripper configuration with conflicting objectives. The primary objective is to find optimum force at the tip of the gripper while grasping the object while satisfying the constraint using the improved multi-objective antlion optimisation algorithm. A comparative analysis has been conducted to evaluate the effectiveness and performance of the proposed algorithm with well-established multi-objective algorithms. The performance analysis for comparing multi-objective algorithm has been carried out using some of the well-recognised performance metrics such as hypervolume, diversity metric, optimiser overhead, and the ratio of non-dominated individuals. From the result obtained after the performance metric study, it is concluded that the multi-criteria design optimisation problem of robotic gripper effectively solved using the proposed I-MOALO algorithm.

中文翻译：

---

一种改进的多目标蚁狮优化算法用于机器人抓手的优化设计

摘要 本文提出了一种进化方法，可以为自动化材料处理过程找到优化设计的夹持器配置。机器人抓手的优化设计是一个非线性、多准则和多约束的问题。引入了进化计算方法来克服在找到夹具的最佳尺寸时相关的困难。在这项研究中，提出了一种改进的多目标蚁狮优化算法（I-MOALO）来寻找三种不同机器人夹持机构的最佳尺寸。多目标优化问题 (MOOP) 是根据具有冲突目标的夹具配置的几何模型制定的。主要目标是使用改进的多目标蚁狮优化算法在满足约束条件的同时，在抓取物体的同时找到抓手尖端的最佳力。已经进行了比较分析，以评估所提出算法与完善的多目标算法的有效性和性能。已经使用一些公认的性能指标（例如超容量、多样性指标、优化器开销和非支配个体的比率）进行了用于比较多目标算法的性能分析。从性能指标研究后获得的结果可以得出结论，使用所提出的 I-MOALO 算法有效解决了机器人抓手的多标准设计优化问题。