A Bio-inspired Network Design Method for Intelligent Transportation
Hanchao Yang, Richard Mayne and Yong Deng
Networks and network models are ubiquitous in modern life. In a transport network, in order to reach peak performance, lowest cost or highest stability, network design is the key issue. More specifically, network design involves selecting edges as a subset of all possible edges, which is regarded as a challenging problem for which universal solutions do not currently exist. In this paper, a bio-inspired method was proposed to address transportation network design. This method adopts a Physarum (slime mould)-inspired algorithm with sequential attractants (food sources) and utilizes the gravity model to predict transportation flux in each city pair. A network adaptation task-based full connection graph with 31 nodes, which was arranged to represent 31 provincial capitals in China, was utilized to test the proposed method. Our results demonstrate enhanced efficiency over similar models and hence demonstrate the viability of the proposed methods for adaptation to problems of transport network optimization.
Keywords: Network design, bio-inspired model, physarum algorithm, gravity model, slime mould