The Research Of Optimal Design Of Urban Rainwater Pipe Networks Based On Reliability

With the development of cities, people pay more attention to the reliability of the drainage network system. However, the existing optimal method of rainwater pipe networks is still using definite value with low reliability, which is unable to meet the needs of urban development. So it is necessary to import the reliability theory to the optimal design of rainwater pipe networks.Considering reliability and optimal design of rainwater pipe networks, the optimal design model of urban rainwater pipe networks based on reliability is put forward in this paper. At first, general design methods of rainwater pipe networks are analyzed and uncertain factors involved in rainwater pipe network system are discussed. The reliability of rainwater pipe networks is calculated to obtain the relationship among the design parameters under a reliable index, which can be used in optimal method of urban rainwater pipe networks as constraint conditions. Secondly, the unit price model of common drainage pipelines are carried out by using regression analysis with genetic algorithms. A more suitable form of the formula and the corresponding local coefficient are achieved, which provide the objective function to optimize the design of drainage channels on computer. In conclusion, the improved genetic algorithm model is established based on the reliability of rainwater pipe network optimization design, through real-coded, proportional selection, the optimal preservation strategy and adding reliability constraints conditions, which is on the basis of basic theory of probability limit state and genetic algorithm optimization method of rainwater pipe networks.It is carried out the optimal design project of diameter and slope under giving the form of arranged pipelines in this paper, using computer program Matlab7.0. Contrasting the calculating result of general design method of rainwater pipe network and of the optimal design model of rainwater pipe network which is be based on reliability theory, ultimately, the theory and applied value of the improved genetic algorithm model are validated.