中文摘要：

固定序算法是Bellman-Ford算法的一种基本改进算法。为了改变固定序算法在稀疏图上的劣势,本文通过预先订制参与迭代的点的计算顺序,对该算法进行了改进。实验表明,在稀疏图上,改进后的算法相对于原算法计算效率提高了近50%,并能够与国际流行的先进先出算法相媲美。本文的工作表明,固定序算法不仅在大规模稠密图上具有明显的优势,而且在稀疏图上也具有很强的竞争力。

英文摘要：

The fixed order algorithm is one of basic improved algorithms on the classic Bellman-Ford algorithm. In view of its inferiority in the sparse directed graph, the algorithm is improved by presenting the computational order of vertices in advance. The experiments show that the improved algorithm is faster than the original one by 50% in the sparse directed graph approximately. Moreover, it is compared with the FIFO algorithm, which is the most attractive basic improved algorithm of Bellman-Ford algorithm at present. It means that the fixed order algo- rithm is very competitive in both the large-scale dense directed graph and the sparse directed graph.