Dissertation > Excellent graduate degree dissertation topics show

The Electrolyte Components’ Prediction Based on Support Vector Machine

Author: TongZhen
Tutor: WangXiaoGang
School: Northeastern University
Course: Control Theory and Control Engineering
Keywords: Soft Measurement Electrolytic copper Support vector regression Multi - objective Genetic Algorithms
CLC: TQ151
Type: Master's thesis
Year: 2008
Downloads: 29
Quote: 0
Read: Download Dissertation

Abstract


Electrolytic production process often involves many complex physical and chemical reactions, interactions and makes the electrolysis process showing a high degree of non-linearity. In order to better control the electrolytic chemical reaction process, it must obtain sufficient electrolytic copper production process information; the thermal electrolytic production process parameters, such as temperature, pressure, flow, level, and so easier to obtain, control also easy to achieve. However, some of the key parameters that determine the quality of products online measurement is difficult to achieve, such as electrolyte copper, acid ion concentration. These parameters are still using laboratory sampling and analysis methods for detection, which tend to produce a large time lag can not be timely feedback data, which makes process control and optimization of electrolytic copper production has been hampered. The soft measuring technique is an effective way to solve this problem. Some information to estimate these are not available online detection of variables, this technology is both easy to implement and cost savings, it is a very viable option. This paper focuses on the soft measurement modeling method based on support vector machines, from support vector machine theory, the choice of the kernel function and parameters were discussed in detail. Comparison of a variety of methods, the selection method for support vector regression model parameters to find a multi-objective optimization parameter selection method, thereby reducing the complexity of the regression model to improve the generalization performance of the model. The problem of over-reliance on the training data for the black box model does not have a physical meaning, article classic electrolysis chemical reaction mechanism of knowledge and support vector machine method proposed a parallel structure of the hybrid model based metal balance equation, volume balance equation create a mechanism model, support vector machine method to compensate for the impact of the mechanism modeling ignores the factors and production status changes on the model, the experiments show that this hybrid model obtained reliable and better generalization performance and reduce dependence on the training data size. Main subject from a soft measurement model structure, the establishment of the the electrolysis reaction mechanism model based on support vector machine modeling methods, the choice of four aspects of the model parameters-depth study of the theory and electrolytic copper to a smelter potroom production research background on the the soft measurement system realization technology simulation study. The approximation effect from soft measurement model to predict the mean square error value, you can see the validity of this method of electrolyte copper acid component prediction.

Related Dissertations

  1. Study on the Soft Measurement for Moisture Content of Acrylonitrile Polymer,TP273
  2. The Research of Robust Image Watermarking Algorithm,TP309.7
  3. Application of Support Vector Regression to Quantitative Structure-activity Relationship,TP18
  4. Shape Prediction Model-Based Face Recognition under Variant Illumination,TP391.41
  5. Based on the research and development of AHP-SVR project risk assessment decision support system,F224
  6. Theoretical and Experimental Investigation of In-Process Measurement of Surface Characteristic Parameter Based on Laser Scattering,TP274
  7. The protein residues depth , flexibility and function prediction and analysis,Q51
  8. Study on the Prediction and Assessment Methods of Water Environment Quality Based on Support Vector Machines Theory,X824
  9. Study on Rational Allocation of Regional Water Resources Based on Sustainable Development,TV213.4
  10. Study on the Volume Fraction Measurement of Oil/Water Two-Phase Flow Based on Conductance Sensor,TP274.4
  11. Standard Entropies of Cations in Solid Compounds with Support Vector Regression,TP181
  12. Regression Analysis and Application of Support Vector Regression in Material Experimental Data,TB302
  13. A Study on the Predicting Model of Gray Neutral Network and Support Vector Machine,TP183
  14. Data Mining Applied in Molecular Phylogeny and Quantitative Structure-Activity Relationship Modeling,Q811.4
  15. Design and Implementation of Neural Network Soft Measurement Platform Based on SOC,TP274
  16. Software Defect Prediction Using Fuzzy Support Vector Regression,TP311.53
  17. Application Study on Short-term Load Forecasting Based on Support Vector Regression,TM715
  18. Cooperative Optimization Scheduling with Application to Multi-Reservoir System During Non-Flood Period,TV697.11
  19. Research on Key Technologies of Network Measurement and Control in Modern Distributed Monitor and Control System and Their Applications,TP277
  20. Study on Integrated Optimal Control System for Aluminum Powder Nitrogen Atomizing & Classifying Process,TP311.52
  21. A Study of EMS Type High Speed Magnetic Levitation Vehicle’s Electromagnetic Systems,U266.4

CLC: > Industrial Technology > Chemical Industry > Electrochemical industries > Electrolysis industry
© 2012 www.DissertationTopic.Net  Mobile