Reliability Design And Neural Network Prediction On Snow Melting Of Ground Source Heat Pump System

During the "Twelfth Five-Year", one trillion yuan will be used for highway construction in China. Although the State attaches great importance to the construction of roads, but snow and ice weather affects traffic, even is the traffic safety archenemy. Especially wide range's sleet weather will cause the wide range paralysis of transportation and will lead huge economic loss. According to the statistics, compared with dry pavement, the road accident is2times when the road is wet, is5times when the road is full of snowfall, is eight times when it freezes. Therefore the pavement snow melting technology is one item which relates to improve transportation condition and people's living conditions, relates to promote economy stable and fast development.Commonly used methods in snow melting are mainly divided into two kinds, they are the passive technology and the initiative technology. The primary passive method is chemical method. It is well known, using of deicing salt will reduce highway's service life, threaten plant's growth on the both sides of road seriously, furthermore will contaminate the soil and the water resources.A new technology of road deicing and snow melting at home and abroad is the thermodynamics method, including conductive concrete, electric heating cables and ground-source heat pump. The ground source heat pump technology makes use of low grade heat energy and heat storage performance of underground shallow soil, and the low grade heat energy can transformed to high grade energy and environmental protection.The ground-source heat pump technology using the solar thermal storage can save energy, and is environmental protection, high efficiency and has great prospects for development. The ground source heat pump snow melting technology is a integrated engineering application system, involving many areas of technology, including road, hvac, geology, mechanics, thermotics, and its application is a comprehensive integration system.The study at home and abroad on the road random unsteady heat transfer research has not made breakthrough, on the asphalt concrete pavement snow melting experimental is still in the blank stage, reliability theory in the snow melting has just started, the above technologies are needed further research. Breakthroughs of ground source heat pump of snow melting will play a role in promoting actual applications in our country.In this paper, according to the present theory of snow melting, a random heat transfer snow melting model is set up based on heat transfer theory.In order to verify correctness and rationality of the theoretical model, indoor experiments are preceded. Heat consumption of the ground source heat pump is predicted by artificial neural network method, and a reliability design method of ground source heat pump snow-melting system is proposed.This dissertation is focus on the following fields.(1)Theory part. The random heat transfer mechanism of asphalt concrete pavement based on snow melting is analyzed and the model of heat transmission is established. The reliability design method and steps of snow melting is proposed on the basis of reliability theory, the reliability equation is established based on the random character of snow melting system. The formula of reliability calculation is deduced which considers the stochastic characters of multiple factors.(2) Test part. The effect of snow melting is researched through the indoor asphalt concrete test under different conditions,including different depth and different spacing of level buried tube,different thermal conductivity of asphalt concrete pavement materials. And heat consumption on different environmental conditions is studied. The results provide practical data for theoretical analysis, but also for the artificial neural network prediction.(3) Numerical experimental research. BP neural network model is established by use of subroutine of MATLAB software, it can be used to predict the snow melting heat consumption. The correctness and rationality of the theory analysis is verified by compared with the results of theoretical analysis and laboratory test.