Intelligent Approaches in Renewable Energy Engineering
智能方法在可再生能源工程中的應用
Student thesis: Doctoral Thesis
Author(s)
Detail(s)
| Awarding Institution | |
|---|---|
| Supervisors/Advisors |
|
| Award date | 26 Jun 2017 |
Link(s)
| Permanent Link | https://scholars.cityu.edu.hk/en/theses/theses(f5544dda-6ff4-4a03-8bea-d7371aa497e8).html |
|---|---|
| Other link(s) | Links |
Abstract
The worldwide pace of the installation capacity of renewable energy including solar and wind has been witnessed along a decade. Such remarkable growth is also expected to continue in the future. Due to the increasing penetration of solar and wind energy in power grids, advanced technologies to smartly accommodate variations of wind and solar power productions are top demanded by power and energy society.
Driven by big data, predictive engineering and optimization with data science offers an emerging tool to deal with the uncertainty and instability in renewable energy conversion. Recent revolutions of data acquisition systems confirm the promising opportunity for generating smarter energy solutions. As optimization problems involving renewable energy systems modeled with data-driven methods are typically non-parametric and non-convex, the improved computational intelligence approaches are widely applied as effective tailor-made solutions for optimization problems containing data-driven elements.
This thesis conducts studies from two aspects to provide data-driven frameworks developed by integrating data mining and computational intelligence approaches for tackling the challenges involved in large penetrations of renewable energy. One aspect of studies targets on the configuration optimization of the renewable energy system to physically improve the overall renewable power production. Another aspect of studies aims at optimizing operational strategies of the renewable energy system to improve the matching between power supply and demand over multiple timescales.
The contributions of this thesis include:
• Model the configuration optimization problem of renewable energy systems. A multi-objective optimization model for optimizing the capacity size of the solar and wind component in a large scale PV/wind system is presented.
• Model the layout optimization problem of renewable energy systems. A comprehensive study of the effectiveness of the classical grid and coordinate wind farm layout models is conducted. A two-echelon wind farm layout planning model is proposed to maximize its expected power output.
• Predict renewable energies with data mining approaches and provide a great support for the operations of renewable power systems.
• Incorporate the data science into the operation optimization problems of renewable energy systems. A two-level model containing data-driven components for studying the robust optimization of operations and power reserves of wind turbines in a wind farm is introduced.
• Solve the non-parametric and non-convex optimization models by computational intelligence. The evolutionary algorithms, such as the random key genetic algorithm and multi-swarm optimization algorithm, are applied to obtain the optimal solutions of the optimization models.
The proposed methods and designed frameworks have been applied and validated with real datasets and optimization problems to demonstrate their advantages and effectiveness.
Driven by big data, predictive engineering and optimization with data science offers an emerging tool to deal with the uncertainty and instability in renewable energy conversion. Recent revolutions of data acquisition systems confirm the promising opportunity for generating smarter energy solutions. As optimization problems involving renewable energy systems modeled with data-driven methods are typically non-parametric and non-convex, the improved computational intelligence approaches are widely applied as effective tailor-made solutions for optimization problems containing data-driven elements.
This thesis conducts studies from two aspects to provide data-driven frameworks developed by integrating data mining and computational intelligence approaches for tackling the challenges involved in large penetrations of renewable energy. One aspect of studies targets on the configuration optimization of the renewable energy system to physically improve the overall renewable power production. Another aspect of studies aims at optimizing operational strategies of the renewable energy system to improve the matching between power supply and demand over multiple timescales.
The contributions of this thesis include:
• Model the configuration optimization problem of renewable energy systems. A multi-objective optimization model for optimizing the capacity size of the solar and wind component in a large scale PV/wind system is presented.
• Model the layout optimization problem of renewable energy systems. A comprehensive study of the effectiveness of the classical grid and coordinate wind farm layout models is conducted. A two-echelon wind farm layout planning model is proposed to maximize its expected power output.
• Predict renewable energies with data mining approaches and provide a great support for the operations of renewable power systems.
• Incorporate the data science into the operation optimization problems of renewable energy systems. A two-level model containing data-driven components for studying the robust optimization of operations and power reserves of wind turbines in a wind farm is introduced.
• Solve the non-parametric and non-convex optimization models by computational intelligence. The evolutionary algorithms, such as the random key genetic algorithm and multi-swarm optimization algorithm, are applied to obtain the optimal solutions of the optimization models.
The proposed methods and designed frameworks have been applied and validated with real datasets and optimization problems to demonstrate their advantages and effectiveness.
