A Solution for Projection Pursuit Problem Based on Particle Swarm Optimization Simulated Annealing

The optimal projection direction for projection pursuit is a class of multivariable constrained optimization problems, with a large solution space. General calculation methods consume much time, are easy to get stuck in local minima, and have low solving efficiency. Simulated annealing can solve projection pursuit problems based on the principle of natural cooling, but there are problems with initial solution setting difficulty and insufficient single round optimization ability. Based on particle swarm optimization and natural population evolution, optimize the initial solution of simulated annealing and promote simulated annealing evolution iteration to improve the efficiency of solving simulated annealing projection pursuit problems. The case results demonstrate the applicability and effectiveness of the proposed method in solving such problems, and the problem solving effect is significantly better than traditional simulated annealing projection search models.