Physarum Polycephalum Inspired Approaches For Road Network Optimization

With the acceleration of urbanization, traffic problems become increasingly prominent. The main reasons to these situations include the unreasonable of road construction and urban planning, the efficiency loss of transportation networks by travelers game and so on. Due to their own complexity, researches focused on the establishment of a low-cost, high-efficiency and reliable networks based on the high transport efficiency in existence have a great theoretical value and practical significance for the development of economy and urban functions.Traffic system is composed of transportation networks, traffic flow and traffic control system. These three parts interact with each other and influence each other, which leading to the operation mechanism of traffic become extremely complex. Hence, a scientific analysis method is extremely demanded. It shall have a comprehensive analysis to the whole transportation networks from a macro perspective, whose final target is to decrease the blindness in road designing and management. At present, people have revealed the topological features, such as mall-world and scale-free of transportation networks. And also have established many models for kinds of traffic elements, including transportation demand, link travel time, the characteristics and behaviors of traffic flow, etc. However, the relationships between travelers game and the modification of network structure, measures about traffic control management still need to be further studied. The two main problems in this field are how to have a scientific plan for road networks and how to have an efficient control to the existing networks.Japanese play a leading role in the researches on traffic problems. It has the most efficient traffic system in the world, and has gained lots of successful experiences. Nevertheless, recent biology studies show that the true slime mold Physarum ploycephalum-a single-cell organism-have the ability to form networks with comparable efficiency, fault tolerance and cost to those of real-world infrastructure networks. Its foraging behavior may provide a useful starting point to guide the planning and designing of infrastructure, such as transportation and communication networks in reality. At the moment, a maze solution model based on the biological mechanism of Physarum have been constructed, it could be applied to searching for the shortest path in networks.This dissertation first studies on the self-organization, self-optimization and self-repair characteristics of Physarum demonstrated in his foraging network. Meanwhile, the detailed process in establishing a complete Physarum mathematical model based on maze model will be talked about. Then, a research on the related elements in transportation network is conducted, at the same time, combining with Physarum model, an algorithm for road networks planning is designed to meet the demands of globalization and intelligence. At last, drawing inspiration from the feedback loops between the thickness of each tube and its cytoplasmic flow in the network optimization of Physarum, the strategy of dynamic optimization for road networks is described in detail.The main content is as follows:1. The maze model is improved.This paper has a plenty of analyses about the optimization process of Physarum network. Meanwhile, the maze model is improved to be more corresponding to actual biological mechanism through introducing the idea of energy. The new Physarum algorithm demonstrates better performance in solving the shortest path problem. Moreover, when expanding to the application in digraph, it can obtain more than one shortest path from one point to several simultaneously.2. Integrated Physarum model is established.This paper establishes a integrated Physarum model to stimulate the network optimization process in the environment of more than one food resources. The model basically takes the advantage of the competition mechanism of flux and energy. It not only can re-appear the process of network optimization, but also maintains the adaptability of network to the changes of environment.3. The biologically-inspired algorithm in road planning is researched.Through analyzing on the differences and common features between transportation networks and Physarum networks, the mathematical models for kinds of traffic elements are established with sufficient consideration of the transportation networks’characters. On this basis, a new optimization policy to Physarum algorithm is proposed. This policy has the similar competition mechanism in constructing the optimization structure and is hoped to have a better guidance in road planning system.4. A strategy of dynamical optimization for road networks is demonstrated.The Physarum networks self-optimize without centralized control or explicit global information by a process of selective reinforcement of preferred routes and simultaneous removal of redundant connections. It is proposed that the dynamical optimization to the whole transportation system can via the control over the road capacity of the existing networks.5. The optimized process is stimulated and analyzed.Based on the above researches, a stimulated platform corresponding with the network optimization is implemented with C# developing language. This platform can display the detailed process of network optimization and dynamical reconfiguration of Physarum networks in visual form. What’s more, it will provide abundant proofs for the model analysis and verification.