Research On ’Black Spot’ Channel Identification And Safety Improvement Strategy In Three Gorges Reservoir Area

After the completion of the Three Gorges Dam,the water depth of the upstream channel of the Three Gorges Dam has been greatly improved,the freight volume has increased rapidly,the draught of navigable ships has deepened significantly,and the overall safety of the channel has been improved.At the same time,the congestion in front of the dam caused by high cargo volume,the reservoir area obstructed by a large number of submerged structures,the increase of landslides on both sides of the bank and the backwater fluctuation area at the tail of the reservoir form a new key waterborne navigation safety control section in the Three Gorges Reservoir Area,which restricts the development of shipping benefits in the Three Gorges Reservoir Area.Scientific identification of the ‘black spots’ channel in the Three Gorges Reservoir Area is of great practical significance for improving the navigation safety of ships in the Three Gorges Reservoir Area and forcefully supporting the implementation of the comprehensive transportation development strategy of the ‘Yangtze River Economic Belt’.‘Black spots’ is generally referred to as the abbreviation of accident-prone sections.Inland water traffic accidents have the characteristics of small samples,data drift,and heavy accident consequences.It has great limitations to define ‘black spots’ only according to the number of accidents.Most of the existing studies focused on the identification of ‘black spots’ from the perspective of post-processing,and the identification results were limited to the section segments of the existing accident records.From the perspective of pre-management,aiming at the channel conditions,this paper studies the ‘black spots’ channel identification and safety promotion strategy selection for the Three Gorges Reservoir Section.Firstly,considering the complexity of navigation conditions and the main influencing factors of water traffic accidents in the Three Gorges Reservoir Area,the connotation of the ‘black spots’ channel in the Three Gorges Reservoir Area was clearly defined from the perspective of identification and former management of the ‘black spots’ channel.Secondly,according to the connotation of ‘black spots’ channel and the characteristics of ship navigation safety in the Three Gorges Reservoir area,the index system of ‘black spots’ channel impact factor was established.One Class Support Vector Machine(OC-SVM)was selected to establish the identification model of ‘black spots’ channel in the Three Gorges Reservoir area based on the channel condition data with accident records.Then,the whole channel of the Three Gorges Reservoir area was identified.The trained model of channel condition data was inputted forward,and the result of channel identification of ‘black spots’ in the Three Gorges Reservoir area is obtained.Finally,the safety promotion strategy and evaluation index of the ‘black spots’ channel in the Three Gorges Reservoir Area were put forward.After qualitative evaluation by expert scoring method,the selection model was established based on Fuzzy Technology for Order Preference by Similarity to an Ideal Solution(Fuzzy-TOPSIS).Considering the weight value of ‘black spots’ channel impact factors and the importance of policy makers to strategic cost and effectiveness under different risk levels,the security promotion strategy was selected for identified ‘black spots’ channel,and the results of strategy selection for single ‘black spots’ channel under different ‘black spots’ channel risk levels are obtained.The identification model of ‘black spots’ channel in the Three Gorges Reservoir Area established in this paper is validated,and the identified ‘black spots’ channel is consistent with the records and key monitoring areas of the maritime department.The model can be used to select different strategies for a single ‘black spots’ channel at different risk levels,which provides a basis for improving navigation safety in the Three Gorges Reservoir area.