Dissertation > Excellent graduate degree dissertation topics show

Further Studies on Methods of Hydrograph Forecasting of Sediment Concentration for Sediment-Laden River

Author: ShiBao
Tutor: QinYi
School: Xi'an University of Technology
Course: Hydrology and Water Resources
Keywords: Sediment concentration forecasting Multiple regression System response function Power system since memorization Neural Networks
CLC: TV143
Type: Master's thesis
Year: 2008
Downloads: 94
Quote: 0
Read: Download Dissertation

Abstract


To reduce siltation of the lower reaches of the Yellow River, and downstream flood control, take a the warping Xiaolangdi sediment regulation and other measures, Xiaobeiganliu, the implementation of these measures to cope with the need to sediment concentration prediction. In this thesis, from the practical point of view, in the the previous sediment concentration forecasting method research based on existing methods improve and further explore the new method of forecasting. The main content of the paper and the results are as follows: 1, multiple regression model is often used to forecast. In its use, the right choice to model the impact of factors affecting the model to predict the effect of the one of the key points. Only consider the existing use of multiple regression model sediment concentration forecasting site traffic and runoff sediment concentration of these two factors. Taking into account the particle size distribution of suspended sediment is an important factor affecting the flood sediment concentration, multiple regression model based on existing, consider suspended sediment particle gradation of flood sediment concentration, established containing The process of sediment load forecasting multiple regression model. The forecast results compare with the impact of suspended sediment particle level and does not consider the multiple regression model, sediment particle size distribution of the multiple regression model to consider the effect of sediment particle size distribution of the multiple regression model forecast, it is better to not consider. Reason, sediment particle size distribution monitoring can not meet the requirements of the forecast, in the model with a large number of the estimated value of the particle size distribution, sediment particle gradation of flood sediment concentration is reflected not only does not come out, but with to a large error is caused. 2, the system response function model is one of the traditional hydrological forecasting method is mainly used in the watershed rainfall runoff forecasting. The view of the evolution of the flood wave in the river has characteristics similar watershed confluence article sediment characteristics based on the lower reaches of the Yellow River for more than a multi-row \Since the system response function model aside grasp of detail, from the macro to the sediment with the evolution of the floods in the river, making the model higher forecast accuracy. Hydrological monitoring status and people's awareness level, the system response function model has a certain practicality and reliability. 3, taking into account the development of things with continuity, also self-memory principle of universality. Based on this power system the first time since the memorization principle apply try to sediment concentration forecasting. Mainly based on the lower reaches of the Yellow River, \Upon examination, the forecast is not ideal. The reason may be to establish the dynamic equations can not effectively reflect the river's sediment characteristics, or that can not be due to the dynamic action which caused the sediment concentration. Dynamic system-memory forecasting model can overcome the shortcomings of the response function forecast lag, it is worthy of further scrutiny of the sediment concentration forecasting pathway. 4, neural networks because of its strong nonlinear mapping ability into many areas of tools to predict things. In order to investigate whether the neural network can be used Tributary river forecast period sediment concentration, as well as input on the impact of network output. Folder riverside station hours sediment concentration prediction neural network model, the input of one of the two programs, one the Huayuankou sediment concentration, flow rate, average flow rate and Jiahetan flow as input factors; Second, the Huayuankou sediment concentration and previous periods Jiahetan sediment concentration as input factor, the output of the network is the folder floodplain sediment concentration. The forecast results show that the input factors between independent network through the learning of the measured data, it will make the role of the input elements cancel each other out and affect the relationship between the input and output, not only can not improve the forecasting accuracy, but reduces the model's forecast effect. 5, by the above four models can be drawn, the system response function model to predict the effect of the best, the two input neural network, followed by, followed by self-memorization model of the power system, the multiple regression model to predict the worst effect.

Related Dissertations

  1. High Speed Frequency Measurment and Non-Linearity Correction of Frequency Modulated Capacitive Displacement Sensor,TH822
  2. Research on Anti-periodic Solutions of Delayed Cellular Neural Networks without Assuming Global Lipschitz Conditions,TP183
  3. Designs and Applications of Fuzzy Synthetic Evaluation Models Based on Parallel Algorithms,TP18
  4. Spectrum Variation of Vegetation in Yanzhou Coal Mine Area and Heavy Metal Stress Characteristic,X173
  5. Multi- license plate location method based on CNN's Intelligent Transportation Systems Research,TP391.41
  6. Research on the Models of GPS Height Fitting Based on BP Neural Network and Their Applications,P228.4
  7. Research of QoS Optimization Based on Neural Network Prediction for Ethernet Passive Optical Network,TN929.1
  8. Sensitivity Study and Application of Complex Variabal Weight Function Neural Network,TP183
  9. Researches on the Application of the Neuron-MOS in the Neural Network,TP183
  10. A Reduction Method for Artificial Neural Network Inputs Based on An Improved Genetic Algorithm,TP18
  11. On-line Monitoring of Tool Wear for Mass Production,TG71
  12. Modeling of Photovoltaic Cells and Its Application in Energy Prediction,TM914.4
  13. Research on Evaluating the Feature Fatigue’s Effect on Customer Equip,F224
  14. Research the Most Important Influencing Factor in a Subjective Evaluation about the Acoustic Environment in Concert Hall,TU112
  15. Delay Neutral System Stability Analysis of Neural Networks,O175.13
  16. One kind of structure in the depth of learning prototype classification method,TP18
  17. The Design of Identification System for Power Marketing Services,TP391.41
  18. The Study of Intelligent Intrusion Detection System Based on Neural Network in Linux,TP393.08
  19. RFID Information Sharing in Supply Chain and Business Valuation Research,F224
  20. Slurry pipeline transportation system, process monitoring and fault diagnosis,F426.1
  21. Consider short-term load forecasting PV,TM715

CLC: > Industrial Technology > Hydraulic Engineering > The basic science of water resources project > Sediment dynamics, River Dynamics > River Dynamics
© 2012 www.DissertationTopic.Net  Mobile