关键词: 冷链配送路径问题, 市区交通, 低碳, 混合蚁群算法

Abstract: In view of the characteristics of frequent changes in traffic speed and many alternative routes in urban areas, the traditional algorithm cannot converge effectively due to the large amount of calculation, a hybrid ant colony and Dijkstra algorithm is proposed. Firstly, Amap API is used to obtain the main traffic roads in the urban area and their speed at different time, and BP neural network is used to predict the speed. On this basis, considering the fixed cost, time variable cost, distance variable cost, time window penalty cost and carbon cost, taking the lowest cost as the objective function, the Dijkstra algorithm with greedy rules is used to search the path, and the road transportation cost is adjusted by constantly adjusting the pheromone left by the ant colony algorithm, a revised cost map is established, the map is called to improve the secondary search speed when the road condition changes, and the algorithm is verified with Python. The example shows that the hybrid algorithm combines the positive feedback characteristics of ant colony algorithm and the strong global search ability of Dijkstra algorithm, shortens the time required to deal with road conditions, and can effectively plan a reasonable path according to the current traffic situation.

Key words: cold chain distribution routing problem, urban traffic, low carbon, hybrid ant colony algorithm