Swarm Robots Search for Multiple Targets Based on an Improved Grouping Strategy.

Area of Science:

• Robotics
• Artificial Intelligence
• Optimization Algorithms

Background:

• Cooperative multi-target search in unknown environments is challenging for swarm robots.
• Existing methods often lack adaptability to dynamic environmental factors and target distributions.

Purpose of the Study:

• To propose an improved grouping strategy for swarm robots to enhance multi-target search efficiency.
• To dynamically group robots based on sensed environmental information and target influence.

Main Methods:

• Developed an improved grouping strategy using constriction factors Particle Swarm Optimization (PSO).
• Robots are grouped after stochastic movements, considering target influence and environmental data.
• Implemented dynamic group structures where each group focuses on a single target.
• Integrated obstacle avoidance mechanisms into the search process.

Main Results:

• Simulations demonstrated superior adaptability, accuracy, and efficiency compared to previous methods.
• The proposed strategy successfully led to the discovery of all targets.
• Dynamic grouping allowed for efficient exploration of unknown environments.

Conclusions:

• The proposed constriction factor-based PSO grouping strategy significantly improves swarm robot performance in multi-target search.
• This approach offers a robust solution for cooperative search missions in complex, unknown environments.