Multi-Agent Cooperative Path Planning with Collision Avoidance Using Hybrid Genetic Algorithm and Neural Networks

Abstract

Multi-agent systems are widely used in various fields, and collision avoidance cooperative path planning has high requirements for
their accuracy, real-time performance, and coordination. To enable multiple artificial agents to avoid collisions and jointly complete certain
tasks in complex environments, a path planning scheme using a hybrid genetic algorithm combined with a neural network is designed. First,
an overall architecture is built and its main requirements and cooperative logic are determined; then, a hybrid genetic algorithm optimization
module is created to improve genetic operators to increase the number of global optimal solutions, and a neural network is used to process the
real-time sensing of collision avoidance information and make collaborative decisions, thereby solving the occurrence of conflict situations;
then, simulation experiments are conducted to verify the advantages of this method in various aspects, achieving the goals of efficient collaborative collision avoidance and path arrangement. The above experimental results show that combining the two algorithms in the path planning
process can improve the coordination and safety of path planning, which has certain engineering application value.