Research Of Channel Congestion Recognition Method Based On Fuzzy Comprehensive Evaluation

Inland navigation as an important part of the transportation system, is not only beneficial to improve the distribution of productivity in river basins, also can promote the comprehensive exploitation and utilization of mineral resources and water. Inland navigation being an environment friendly transport way can meet the demand of material transportation and other various aspects. But in recent years increasing channel congestion events gradually become the "bottleneck" of the development of ports and threat the safety of transport, not only seriously affecting the sustainable development of Inland navigation, but also bringing huge pressure to other transportation industries. This paper uses the fractal theory to investigate the complexity of the vessel traffic flow dynamic system, and predicts the congestion phenomenon with the ship’s GPS data and the information of waterway. This method can help the relevant departments make the reasonable decision.In chapter one, the research on the channel is mainly concentrated in the waterway capacity, channel security level. Whether in theory or in practical application the study on channel has been mature developed. However, rare studies are found about the channel congestion problem, consequently it deserves putting effort on this problem.In chapter two, we analyze factors that influence the channel capacity such as channel depth, width and curve. Through the study on congestion evaluation method and congestion analysis we define congestion and give method to evaluate it.In chapter three, we use the fractal analysis method to study the self-similarity characteristics in channel transportation system. The result shows that the fractal dimension of vessel traffic flow varies great on different state. Then with the rescaled range analysis method we calculate the long-range correlation and memory length. The result shows that under three different conditions, the time series of vessel traffic flow is positive related long memory. However, the memory cycle is not stable, and this demonstrates that the complexity of the traffic flow system varies under different conditions.Finally, we analyze the vessel traffic flow data and the commerce data, and then construct the congestion evaluation model based on fuzzy comprehensive evaluation method. We use the GPS data to validate this model, and the result shows that fuzzy comprehensive evaluation method performs well in identifying inland channel congestion.