Optimized Control of Grid-connected Photovoltaic System

摘要

Increasing environmental concerns regarding the inefficient use of electrical energy fromfossil oil have directed attention to the importance of producing electric power from renewableenergy resources such as solar and wind energies which are environmental friendly, and on topof that it has no gas emissions.Solar energy plays a major role since it is globally available and itis flexible with regard to the system size and because it can fulfill the needs of different countriessince it offers on-grid and off-grid solutions.The use of PV systems in electricity generationstarted in the 1970s and today is growing rapidly around worldwide in spite of high capital cost.
　　The performance of a PV system largely depends on solar radiation, temperature and conversionefficiency.Although PV systems have advantages, but they have suffered from weathervariations, high installation cost, and low efficiency that is less than 20％ for module.Therefore,the predication of PV system is an important aspect to describe the performance with differentweather conditions.
　　PV predication has become very interesting in electrical applications.This is due in largepart as the source of widely available in some areas.The major drawback is the fact that theoutput power generated by PV panels depends on weather conditions (solar radiation andtemperature).To improve this problem and in order to maximize the power, it is necessary topredicate solar power to track the maximum power point of the PVPG.PV has a single operatingpoint that can provide maximum power to the load.
　　In this thesis the solar system optimization and control has been studied and thepredication has been explained using the Forgetting Mechanism Online Order Extreme LearningMechanism (FOS-ELM), which is the most recent study about solar power predication.Powerpredication based on (FOS-ELM) theory is proposed.Presented power optimization controlscheme not only has the ability to prevent the network voltage limit, but it also makes full use ofthe PVPG system's own function, which is conducive to the regulation of energy consumptionand output of large scale PV systems.Influence by access the great number of the PVPG systemshas been discussed.Small PVPG system analysis has been presented.According to which, whenthe distribution active power is too high it will lead to conclusion that low voltage feeder islimited.To get the maximum PV power out a scheme to solve the overpressure problem of thefeeder is been proposed.Optimal power flow (OPF) is combined with the sensitivity matrixanalysis to optimize maximum power of each photovoltaic.With the proposed control system, itis proved that the proposed dual power optimization control is more feasible.The MATLABsimulation software is used to analyze the examples.The results show that the PV powerprediction method can help to optimize the operation of the GCPV system.

目录

声明

ACKNOWLEDGEMENTS

ABSTRACT

INITIALISMS

Table of Contents

Chapter 1 Introduction

1．1 Overview of photovoltaie power generation

1．2 PV power generation development

1．3 Key technologies used in GCPV system

1．3．1 Optimal planning of PV systems

1．3．2 PV power prediction

1．3．3 PV power stability and its control technology

1．4 Main research content

Chapter 2 Photovoltaic Power System

2．1 Introduction

2．2 Photovoltaic power generation(PVPG)system

2．2．1 Photovoltaic cell

2．2．3 PV converters

2．2．3 PV inverters

2．3 Grid-connected PV inverters

2．3．1 H-bridge Inverter with Uni-polar and Bipolar Modulation

2．3．2 HERIC Inverter

2．3．3 H-5 Inverter

2．3．4 Comparison

2．4 Inverter relation with maximum power point tracking(MPPT)

2．5 Summary

Chapter 3 Photovoltaic Power Predication

3．1 Impact of PVPG on network

3．2 Online Sequential Extreme Learning Machine with Forgetting Mechanism(FOS-ELM)

3．3 Prediction Algorithm

3．3．1 Extreme Learning Machine(ELM)

3．3．2 Online Sequential ELM(OS-ELM)

3．3．3 OS-ELM with Forgetting Mechanism(FOS-ELM)

3．4 Model architecture

3．5 Simulation

3．6 Summary

Chapter 4 Grid-connected PV Optimization Control

4．1 Optimum power Distribution of grid-connected PV

4．2 Power Optimization PV Generation System

4．3 Example simulation verification

4．3．1 Conventional use of sag power optimization control

4．3．2 Sensitivity Matrix of Power Optimization

4．4 On-site acquisition test

4．5 Summary

Chapter 5 Conclusion

References