In forest, we planed roads for recreation with a digital map and studied the most suitable route of them using a genetic algorithm (GA) and Dijkstra method to do total evaluation of them efficiently. Considering earth-volume and safety of driving, we confined them to one-way traffic and one-lane road in this paper. We planed two roads to go and back in a different route. The evaluation factors are the scene in a forest, the view, the viewed frequency, the slope gradient, the distance between two roads, the variance of ground height, the length of roads and the longitudinal slope. One calculation of this method is to settle two intermediate points per one route, and to search between four points, beginning point, two intermediate points, and end point, in Dijkstra method. Changing coordinates of the intermediate point, we attempt a solution of the most suitable one by a simulated evolution of GA. We obtained the optimum solution, calculating in 30,000 times. It took about 45 hours. We can consider the factors that are evaluated with looking out the whole area and do total evaluation. And we can confirm the efficiency of GA. The future problems are studying how to select the factors and weights that justly fit the size of the grid cell and settle intermediate points.
View full abstract