Fuzzy Clustering with Obstructed Distance Based on Quantum-Behaved Particle Swarm Optimization

Typical partitioning methods of constraint-based spatial clustering algorithms are based on gradient descent, which are easily falling into local extremum and sensitive to the initial parameters. A new fuzzy clustering with detour distance algorithm based on quantum-behaved particle swarm optimization (QFCOD) was proposed. The new spatial clustering with obstacles constrained algorithm avoids the fitness value of clustering falling into local extremum in a large degree. Furthermore, QFCOD adopts membership grade in the object function of QPSO, redefines detour distance and applies the Particles Escaping Principle to avoiding that the updated cluster center particle sinking into the area of the obstacles. Finally, this algorithm illustrates effectiveness and accuracy on the basis of the experiments running in the Matlab environment and the sample points with obstacles.