An Area Exploration Algorithm With Swarm Robots Using Cellular Automata Based on Multi-information Superposition.

This article proposes a cellular automata multi-information superposition model for swarm exploration in intricate, unfamiliar environments. The core novelty lies in its decision-making logic, which fundamentally departs from traditional cellular automata models that rely on a single environmental signal, such as pheromones. The model explicitly decouples and adaptively weighs multiple information streams-the number of local robots, distance, and repeated times of exploration-to guide each agent. This multifaceted guidance mechanism allows the swarm to avoid the local optima traps inherent in single-signal systems. The model's effectiveness is numerically demonstrated through comparative simulations against established benchmarks, including a pheromone-based repulsive model and a biased random walk model. Results indicate that the model achieves a significant improvement in exploration coverage rate and reduces redundant visits, demonstrating superior efficiency. The proposed model demonstrates generalizability across multiple enclosed environments with complex boundaries. Simulations across these diverse structures show consistent coverage completeness, with exploration efficiency varying by environmental configuration. Though the model accelerates exploration in most tested scenarios, certain topologies, such as annular structures, present opportunities for further efficiency optimization.