Particle Swarm Optimization Algorithm Based on Graph Knowledge Transfer for Geometric Constraint Solving

In order to more effectively solve complicated geometric constraint problems by applying swarm intelligence technologies, a particle swarm optimization (PSO) algorithm based on graph knowledge transfer for geometric constraint solving (GCS) is proposed. By fusing the graph knowledge transfer mechanism into the PSO algorithm to select parameters deciding the algorithm performance, avoiding getting stuck in a local extreme value and then making the algorithm stagnating when solving a practical complicated geometric constraint problem. Empirical results show that using the graph knowledge transfer mechanism to select the parameters of PSO can obtain high-quality parameters of GCS. It improves the efficiency and reliability of GCS and possess better convergence property.