Multiple route planning algorithm based on improved K-means clustering and particle swarm optimization

To improve the survival rate of aerial vehicle in battlefield, a method that provides multiple alternative routes for it to choose and replace is proposed. For this problem, threat models of aerial vehicles are built to generate the basic cost functions of route planning. Then, a new strategy named exclusion mechanism is introduced to improve K-means clustering, which improves the variety of solutions and contributes to high routes' spatial dispersion. Thanks to the improved K-means clustering, the routes can be classified owing to their distribution in space. Finally, to enhance the efficiency of solving, particle swarm optimization(PSO) is chosen to make the algorithm adaptable. The simulation compares the proposed algorithm with a related one, which proves that, unaffected by subjectivity of artificial planning, the improved algorithm can finish multiple route planning quickly and meet the demand of pre-route-planning in actual combat.