Font Size: a A A

Application And Study On Water Demand Mixed Forecasting Model Of Urban

Posted on:2007-12-14Degree:MasterType:Thesis
Country:ChinaCandidate:M W WangFull Text:PDF
GTID:2132360212966823Subject:Municipal engineering
Abstract/Summary:PDF Full Text Request
Forecast of water consumption in urban areas could be divided into such catalogues as: medium and long-term yearly water consumption forecast, short-term hourly, daily and monthly water consumption forecast. They are not only the effective means of managing and planning the municipal water resource, but also the important precondition of optimal control of the municipal water distribution network. At present, single water demand forecasting model is adopted universally across the world. However, in light of the complicated influent factors of water consumption, satisfactory forecasting precision isn't obtained sometimes through making use of single water demand forecasting method. So, a combination of different models becomes more attentive in forecasting water consumption.The various basic means about forecast of water demand in urban areas is summarized systematically in this paper. Based on these methods, corresponding mixed water consumption forecasting model is put forward with combination of the practical project compare and different water consumption models.On the basis of the variable characteristics of urban monthly water consumption, it is necessary to select reasonably the equal dimension and new information data that represents the variable regularity of urban monthly water consumption and output expectancy to establish an equal dimension and new information neural network model of urban monthly water consumption. By a practice in a certain city and in comparison with other methods, it is proved that the forecasting error is so little to meet the practical requirement of planning water resource.Taking Jiamusi city's daily water consumption for example, a new urban daily water consumption model has been established based on the strongpoint of genetic algorithm and the BP algorithm. Through multi-group data testing it and being compared to BP algorithm, the results show that GA-BP is an effective and accurate method to forecast urban daily water consumption.It is very significant to study the variable regularity of urban water consumer's water consumption for commanding the variable regularity of water...
Keywords/Search Tags:water consumption forecast, BP neural network, GA-BP mixed method, SE-GM mixed method, dynamic water consumption forecasting model
PDF Full Text Request
Related items