Development of an improved Jellyfish Search (JS) algorithm for solving the optimal path problem of multi-robot collaborative multi-tasking in complex vertical farms

This paper proposes an improved Jellyfish Search algorithm, namely TLDW-JS, for solving the problem of optimal path planning of multi-robot collaboration in the multi-tasking of complex vertical farming environments. Vertical farming is an efficient way to solve the global food problem, but how to deploy agricultural robots in the environment constitutes a great challenge, which involves energy consumption and task efficiency. The most important improvements introduced by the proposed TLDW-JS algorithm are as follows: the Tent Chaos used to generate a high-quality, diversified initial population, Lévy flight used in the improved JS to strengthen global exploration, and finally, the nonlinear dynamically weighted adjustment with logistic functions to balance exploration and exploitation. A Vertical Farming System Multi-Robot Collaborative Trajectory Planning (VFSMRCTP) model has been developed in accordance with the environmental constraints specific to vertical farms, the task constraints, and the constraints between agricultural robots. The VFSMRCTP model is solved using the TLDW-JS algorithm and a number of comparison algorithms in order to analyze the algorithm’s performance. Comparative experiments demonstrate that TLDW-JS outperforms classic optimization algorithms such as the Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Dung Beetle Optimization (DBO), achieving superior path length optimization, reduced energy consumption, and improved convergence speed. The results indicate that TLDW-JS achieved a 34.3% reduction in average path length, obtained one of the top three optimal solutions in 74% of cases, and reached convergence within an average of 55.9 iterations. These results validate the efficiency of TLDW-JS in enhancing energy optimization and demonstrate its potential for enabling automated systems in vertical farming.

Text 121863.pdf - Published Version Available under License Creative Commons Attribution. Download (8MB) Official URL or Download Paper: https://www.mdpi.com/2077-0472/15/6/578

Actions (login required)

View Item