Multi-Strategy Improved Dung Beetle Optimization Algorithm for Multi-Robot Task Allocation

Task allocation in multi-robot systems (MRTA) is a critical issue that needs to be addressed. The emerging Dung Beetle Optimization (DBO) algorithm successfully integrates global extensive searching and local fine-tuning capabilities through its flexible search mechanism. In an effort to optimize the capabilities and investigate the applicability of DBO in MRTA scenarios, we developed an advanced variant called Multi-Strategy Improved Dung Beetle Optimization (IDBO). The initialization phase incorporates a chaos-based mapping technique, which is used to create an initial population with more uniform distribution and enhanced diversity. To achieve an optimal balance between exploration and exploitation capabilities, the algorithm implements a nonlinear convergence coefficient when updating the locations of both breeding and small dung beetles. A simulated annealing strategy is implemented in the position updating of theft beetles, which helps prevent the algorithm from falling into local optima. Finally, the improved algorithm, along with the DBO and Genetic Algorithm (GA), is applied to MRTA in a warehouse environment for simulation experiments. The results demonstrate that the IDBO algorithm achieves higher and faster convergence accuracy and has a better capability to escape local optima, showing superior performance in MRTA problems.