摘要
正确的仓库选址,不仅能够提高运输效率,还能降低运输成本,对货物运输有着重要的作用。本文通过初始化蚁群算法距离矩阵,迭代计算不可直达客户点之间的最短距离,实现对蚁群算法的改进;相较于传统的DBSCAN聚类算法,我们将最短距离矩阵作为算法的输入,提出了基于Distance自适应的DBSCAN算法(Distance_DBSCAN),对核心点的选择策略进行修改,实现对DBSCAN算法的改进;针对聚类结果再次利用蚁群算法,给出每一个类别的仓库选址位置和最短路径策略。实验表明,本文提出的Distance_DBSCAN算法可以正确分离噪声点,有着较好的ARI指数,且时间复杂度低,可以有效缩短聚类总路程,更重要的是此算法更加符合货物运输仓库选址的实际意义。
The correct location of the warehouse can not only improve the transportation efficiency, but also reduce the transportation cost, which has an important role in the transportation of goods. By ini-tializing the distance matrix of ant colony algorithm and iteratively calculating the shortest distance between inaccessible customer points, this paper realizes the improvement of ant colony algorithm. Compared with the traditional DBSCAN clustering algorithm, we take the shortest Distance matrix as the input of the algorithm, and propose an Adaptive DBSCAN algorithm based on distance (Dis-tance_DBSCAN), modify the core point selection strategy, and realize the improvement of DBSCAN algorithm. According to the clustering results, the ant colony algorithm is used again to give the warehouse location and the shortest path strategy for each category. Experiments show that the Distance_DBSCAN algorithm proposed in this paper can correctly separate noise points, have a good ARI index and low time complexity, and can effectively shorten the total clustering distance. More importantly, this algorithm is more in line with the practical significance of cargo transportation warehouse location.
出处
《应用数学进展》
2023年第12期5027-5038,共12页
Advances in Applied Mathematics
