The Model Test Of Ice-block Cable And The Predict Of The Ice-Blocked Thickness Of Transportation Channels

The South-to-North Water Transfer Project is the world's most magnificent inter-basin water transfer project. The project will effectively resolve the serious shortage of water resources in north China.,establishing the reasonable allocation of water resources, promote the pattern of economic and social and ecological environment, harmonious development. Such problems as drift ice, ice jam and ice dam always appear in winter, affects the normal water delivery and construction safety. To guarantee the project security must solve the problems below: research and optimize the control measures of ice disaster, prevent and mitigate the ice disaster。This research put forward a new kind of effective ice-block cable structure with dual-shaft. Based on the model tests for ice-block cable, this paper not only introduces the experimental equipment, model design and processing, test principle, test procedure detailedly, but also analysis the the ice-block cable's working principle, failure mode and so on.Through the observation of model test and analysis of the test data, this article systematically concludes the ice-block cable's process and failure mode and the factors: water level, ice velocity, the initial ice thickness which effect the ice-block cable, include the block of ice thickness, failure mode and the shaft force's trends in different conditions.Combined with actual ice condition, analyzed the model test results, calculated the maximum of ice-blocked, further, calculated the relationship between the effective distance of each ice-block cable and the effect factor as channel width, ice thickness and ice velocity.Finally, based on the BP algorithm and genetic algorithm to optimize the BP neural network establish model test of measured data, predict the thickness of each section before the the ice-block cable in different conditions. Comparative analysis shows that the errors of forecast data are most in reasonable range, and theGenetie Algorithm optimization neural have more advantage, the prediction and measured curve fits well. This research can provide beneficial references to the forecast model for the ice control measures to the South-to-North Water Transfer Project.