Application Of Parking Based On An Improved PSO Algorithm

Particle swarm algorithm, which simulates the foraging behavior of the bird population, achieves the best by the cooperation of birds. Although the behavior rule of individuals is simple, the behavior of the combined population is very complex. This algorithm is based on interating and the population searches in solution space followed by the optimal swarm. Easy to implement and profound intelligent background is the algorithm's advantages. It is fit for both scientific research and engineering application.Based on the researches on the basic principle and research actuality and considering PSO algorithm has several problems, a novel PSO algorithm is proposed. And it is applied in parking guidance. The main research works are as follows:1,Dynamic proportional factor is introduced into the differential evolution algorithm and crossover and mutation operators are introduced into PSO algorithm to rebuild the new location update formula and propose a novel DE-PSO algorithm.2,Before the particles evolving to the next generation, crossover and mutation operators from the GA is selected to optimize the particles. The new algorithm ensures the powerful global searching ability of the GA and also merges the location transfer method of the PSO algorithm. The new algorithm structures a novel CMPSO algorithm by utilizing the information of both the population and the individuals, which is ignored by GA, and the idea of'Survival of the fittest'.3,The traditional PSO algorithm adopts inertia weight adjustive strategy of linear decrease generally. This strategy always disables the PSO algorithm from showing the process of non-linear optimized searching. And because of the linear decrease of inertia weight, particles easily fall into"prematurity"in the later phase of convergence. Considering this problem, difference factor is introduced and the inertia weight of PSO is adjusted dynamically.4,How to design the fitness function is the difficulty and emphasis of the problem of applying the PSO algorithm to parking guidance. This paper utilizes neural network to construct fitness function, which depicts the environmental restriction and distance information of the path.5,The improved PSO algorithm is applied to solve the problem of parking guidance and the simulation results were provided to verify the effectiveness and practicability of this approach.