Gis Path To Find The Best Of The Ant Colony Algorithm

The shortest path optimization is one of the critical problems in transportation networks. It's also the basis of many areas, such as resource allocation and route designing. Over the years, development, application, and efficiency analysis of the shortest path algorithm have been a hot spot in operational research, geographic information science and so on. As a result, large numbers of algorithms came out. And these algorithms are mainly focused on the researches and applications of improvement based on classical algorithm-Dijkstra algorithm. Ant colony algorithm developed rapidly in recent years, it is a global optimization algorithm and based on the principles of biological evolution. However, this algorithm has not yet formed a systemic and mature theoretical system. To improve searching speed and overcome the premature convergence is still the direction of the algorithm. The purpose of paper is improving efficiency of the algorithm and studying searching strategy of ACO based on road networks. And paper also discussed the shortest path problem in time-dependent networks. The main tasks are following aspects:1. Through the extraction of roads from electronic map of Taizhou City, and construction of topology network, to construct the road network platform for the shortest path analysis.2. Based on the platform above, design and implement ACO in optimization of geographical roads, including: analysing process of ant searching, designing steps of the algorithm, coding and implementing.3. After implement ACO in optimization of geographical roads, use the strategy of searching region limited to reduce the searching region when ants search. Experimental results show that: limiting the searching region, the number of nodes searched by ants is less, and the time is shorter.4. Using ACO's multiple-solutions, and by adjusting the pheromone release standards, to ensure the quality of ant sub-optimal solution, results show that the shortest path is not feasible, the ants will more quickly converge to the path up the best alternative.5. Based on the analysis of time-dependent networks, construct a model of time-dependent road networks. Based on the data, get from real road network, by streamline and speed, simplify time-dependent networks. In the model, implementing time-varying shortest path algorithm.