Slime Mould Algorithm using Latin Hypercube Sampling and Improved Strategy for Solving Robot Gripper Problem

Robotic grippers play a key role in various industries, including manufacturing, logistics, and healthcare, such as enabling the robot to grasp, hold, and manipulate objects. It is essential to develop and design the best configuration for the robot gripper. In this paper, the robot gripper’s geometric configuration is optimized through proposing a novel variant of Slime Mould Algorithm (SMA). The SMA is one of the metaheuristic methods that mimics the slime mould’s foraging behavior. To improve the SMA, a novel algorithm named, LHISMA, is introduced to generate the optimal outcomes while taking into account its shortcomings and features. In order to improve global search performance and attain a more uniform distribution throughout the search space, Latin Hypercube Sampling (LHS) is used to initialize the population with the goal of covering as much of the solution space as possible. An improved search strategy is introduced in SMA to enrich the population’s diversity, strengthening its exploitation ability and enhancing the proposed algorithm’s convergence accuracy. Experiments employing 13 benchmark functions, encompassing both unimodal and multimodal types, and the CEC2022 test suite were conducted to evaluate the proposed algorithm’s performance. The experimental findings indicate that the LHISMA algorithm achieves superior convergence accuracy, faster convergence speed, and enhanced global search capabilities compared to other methods. The complex robot gripper problem is solved with the LHISMA algorithm, and the results are compared with the most popular algorithms. The findings from the comparison show that the LHISMA is better than its competitors. The significance of the results is also examined statistically using the Wilcoxon rank-sum test, the Friedman test, and post hoc statistical tests. The findings show the advantages of the LHISMA algorithm in terms of solution accuracy, and the convergence curve and boxplot illustrate the effects of algorithms used to optimize the robot gripper problem.