Particle Swarm Optimization Algorithm In Path Planning Applications

Urban transport affects the development of major cities and cultural technology and tourism. People want a city where traffic facilities, reasonable traffic planning, which will facilitate travel. So the path planning of urban becomes particularly important. According to travelers'purpose and travel requirements, the path planning problem can be specific into the Traveling Salesman Problem (TSP), such as:path around the city attractions. The problem is solved by the path search algorithm. Path search algorithm develops from the traditional search algorithm to the intelligent search algorithm, because intelligent algorithms show good performance in recent years, such as: Genetic Algorithm, Ant Colony Algorithm, PSO Algorithm (Particle Swarm Optimization). PSO is the simple concept, easy to implement, better robustness and so on.Currently, the solving process is so complex as to solve slowly when the PSO is applied to the TSP. Results can not show advantage of overall performance, and experiment mostly studies simulation data. Therefore, this paper studies the actual geographical data to solve the TSP, and details a discrete PSO algorithm model based on geographic coordinates for solving the practical TSP. To further improve the global search capability of the algorithm, the paper provides a self-balancing mechanism to make the process simple and improve algorithm performance. Meanwhile, the paper uses the multi-threaded parallel to improve the speed of algorithm and makes up the lack of parallel algorithm research. This paper has four major aspects:(1) PSO algorithm solves the TSP in the actual city attractions. There is a discrete PSO algorithm based on the geographical coordinates for TSP application. Experiments calculated and recorded the optimal solution and the average value of the same iteration's the different number of particles'this algorithm. Then, the optimal solution and the average value of the different number of iteration's algorithm are statistically compared. Results show that the algorithm has important practical significance for solving practical problem, but the algorithm is so easy to fall into local optimization.(2) In order to avoid local optimization and make the process simple, there is a self-balancing discrete PSO algorithm for TSP application. Compared for other algorithms, this algorithm balances search ability by itself. Experimental statistical methods are same as a discrete PSO algorithm based on the geographical coordinates, and results show that solving value is better than the former. It makes the process simple. The algorithm can effectively balance the PSO search ability. (3) In order to further improve the efficiency of a self-balancing discrete PSO algorithm, the paper provides the parallel model. It uses parallel multi-threaded OpenMP approach based on the multi-core platforms to improve the algorithm performance. Because of a lack of parallel studies of this algorithm, this parallel model not only promotes the development of parallel algorithms, but also shows that it can effectively improve the algorithm performance by experiment and results.(4) This paper designs and completes the path planning application system of these PSO algoritlims. The system is constituted by the practical map data, parameter setting module, algorithm modules and results module. Experimental results show that the performance of PSO algorithm is gradually improved, and it can effectively solve the practical TSP. At the same time, it can provide path reference of construction in Xi'an Metro departments to promote development of the city's attractions around the track, urban transportation and economy.