Research On Emergency Material Distribution Route Based On Remote Sensing Image Road Information

In recent years,our country has experienced earthquakes frequently.After the earthquake,the disaster areas are in urgent need of a great deal of relief supplies.The disaster areas are often accompanied by the paralysis of basic communication facilities,and the disaster relief command center can not obtain the information of damaged roads in time.There is a lot of uncertainty in the distribution of relief supplies,which seriously restricts the timeliness and effectiveness of disaster relief.Timely and effective distribution need to be based on accurate road damage information to make a reasonable path planning program after the earthquake.This paper takes the relief supplies vehicle path planning as the research object,proposes an extraction method based on the disaster damage situation of high-resolution remote sensing imagery in the disaster area,assesses its traffic capacity according to the road damage situation,and designs an algorithm to achieve road emergency repairing and relief supplies distribution with the planning of the route.The main research contents include:(1)Extracting the debris flow area on the road after the earthquake.A comprehensive judgment method combining the detection of road edge integrity and debris flow area extraction is proposed.By using the improved Canny edge detection algorithm to detect image edge information,an adaptive threshold Hough detection algorithm in ?-?domain is proposed to implement road edge detection and count the number and position of edge interruptions.Extract the positive and negative samples of a large number of debris flow image regions,and use a linear kernel SVM classifier to extract the debris flow target region.In combination with the location of the debris flow area and the location of the road line interruption,the location and number of road segments damaged by the debris flow are determined to provide a basis for the subsequent road damage models.(2)Design and verification of path planning algorithm.Proposed variable road block model and emergency logistics route planning model.The improved genetic-ant colony(GA-ACO)algorithm is used to optimize the designed model.The unbalance fitness function f based on the number of generations is introduced to solve the premature phenomenon in the optimization of genetic algorithm.The pheromone volatile factor ?of fast fading based on the Newtonian cooling model is proposed so that the near optimal solution can be quickly picked out to speed up the convergence speed of the algorithm.The simulation verification algorithm has certain practical significance for the post-earthquake emergency logistics route planning.(3)Critical road repairing and relief supplies distribution optimization.Aiming at the characteristics of the post-earthquake road network,a dual optimization model of road repair and relief supplies distribution is established.A variable neighborhood search algorithm based on distance matrix is proposed to solve the model.By designing the neighborhood structure,the problem of catching in the local optimal is solved.