Dissertation > Excellent graduate degree dissertation topics show

Neural Network PI Pepetitive Controller for Active Power Filter

Author: HuangZuo
Tutor: ZhengYiHui
School: Shanghai Jiaotong University
Course: Theory and New
Keywords: Active Power Filter Repetitive control Neural Networks PI control Recursive integral PI control Decoupling Control
CLC: TN713.8
Type: Master's thesis
Year: 2011
Downloads: 38
Quote: 0
Read: Download Dissertation

Abstract


In recent years, with the rapid development of power electronics technology, electric power systems have emerged more serious power quality problems, harmonic pollution, reactive power, voltage fluctuations and unbalance. Active Power Filter (Active Power Filter, APF) is recognized as the governance the grid harmonic and reactive pollution, effective means of improving the power quality, has become a new research focus in the application of power electronics technology. But the APF in domestic applications is far from mature, compared with the passive filter, there are many issues to be further research and improvement. APF by generating harmonic currents having the same amplitude and opposite phase of the compensation current to achieve the purpose of the elimination of harmonics, so the design of the controller on the output of the active filter can quickly and accurately tracking the reference current play a key role. In order to obtain the desired compensation characteristics, taking into account the active power filter system instruction signal for the characteristics of the periodic signal for repetitive controller cause a cycle delay defects designed neural network PI repeat controller. The controller uses a neural network to optimize the PI parameters to improve the dynamic response of the system, making the APF command current track real-time changes quickly to make up for the repetitive controller delay defects; duplicate controller to improve the harmonic tracking when the system enters the steady state, accuracy, eliminate the steady state error. Neural network Repeat PI controller to optimize each sampling time are subject to the neural network learning parameters computationally intensive, and work in parallel recursive integral PI controller is equivalent to N PI controller, just in a cycle neural network learning, can greatly improve the rapidity and accuracy. Therefore, we design a neural network recursive integral PI repeat controller, recursive integral PI controller parameters online tuning through the neural network, to speed up the response speed recursive integral PI control, while taking advantage of the repetitive controller to improve tracking stability state accuracy, therefore has a small steady state error, computing fast, easy engineering advantages, compared with previously proposed neural network PI repetition controller, has significantly improved in the APF in the initial tracking speed and accuracy of the initial tracking. However, the above method only applies to the grid is completely symmetrical, in engineering practice, the three-phase power grid has negative sequence current asymmetry due to load this recursive integral PI proposed neural network decoupling of the positive and negative sequence repeat controller. Independent control of the controller of the positive and negative sequence current for positive and negative sequence current tracking no static error, 2 octave harmonic suppression grid imbalances caused by the APF DC side voltage inverter circuit, in order to achieve the grid the asymmetric situation harmonic tracking, improve the overall performance of the system. Finally, the simulation analysis proved the validity of the above algorithm.

Related Dissertations

  1. Research on the 6-Dof Fault Tolerant Control of the Vibration Isolation Platform with Eight Actuators,TB535.1
  2. High Speed Frequency Measurment and Non-Linearity Correction of Frequency Modulated Capacitive Displacement Sensor,TH822
  3. Study of Three-Phase AC Chopper Surge and Its Harmonic Elimination Technology,TM423
  4. Research on Anti-periodic Solutions of Delayed Cellular Neural Networks without Assuming Global Lipschitz Conditions,TP183
  5. Designs and Applications of Fuzzy Synthetic Evaluation Models Based on Parallel Algorithms,TP18
  6. Spectrum Variation of Vegetation in Yanzhou Coal Mine Area and Heavy Metal Stress Characteristic,X173
  7. A Novel Controller Design for Active Power Filter with Low Switching Loss,TN713.8
  8. Multi- license plate location method based on CNN's Intelligent Transportation Systems Research,TP391.41
  9. Simulation System for Aluminum Foil Cold Mill and Research of Tension Control,TG334.9
  10. Digital Delta Type UPS Design and Research,TN86
  11. Research on the Models of GPS Height Fitting Based on BP Neural Network and Their Applications,P228.4
  12. The Design and Study of Control System of Car-type Heat Treating Furnace,TG155.92
  13. Research of QoS Optimization Based on Neural Network Prediction for Ethernet Passive Optical Network,TN929.1
  14. Sensitivity Study and Application of Complex Variabal Weight Function Neural Network,TP183
  15. Researches on the Application of the Neuron-MOS in the Neural Network,TP183
  16. A Reduction Method for Artificial Neural Network Inputs Based on An Improved Genetic Algorithm,TP18
  17. On-line Monitoring of Tool Wear for Mass Production,TG71
  18. Modeling of Photovoltaic Cells and Its Application in Energy Prediction,TM914.4
  19. Research of Active Power Filter in Three-phase Four-wire System,TN713.8
  20. A Shunt Active Power Filter with Method of Hybrid Control,TN713.8
  21. Research on Evaluating the Feature Fatigue’s Effect on Customer Equip,F224

CLC: > Industrial Technology > Radio electronics, telecommunications technology > Basic electronic circuits > Filtering techniques,the filter > Active filter
© 2012 www.DissertationTopic.Net  Mobile