Energy management and controlling of microgrid using improved and enhanced model predictive control

Abstract

The widespread adoption of renewable energy sources (RESs) has occurred in an attempt to stop the progression of global warming. This growing adaptation has resulted in a significant shift in the topologies of traditional power networks, and as a result, the concept of a microgrid (MG) has emerged. MGs represent a paradigm shift from remote central power plants to more localized distributed generation and are a growing sector of the energy industry. Controlling MGs is difficult due to their complexity and diverse properties each asset in the MG has. Various solutions have been proposed to address the difficult problem of MG management. Some of these methods are considered optimal for managing MG assets. Other works are based on a systems-based methodology and address the scalability and simplicity of synthesizing an energy management system (EMS) for an MG. MPC is a sophisticated method used to control power systems while satisfying multiple constraints to achieve an optimal solution based on various criteria. MPC is an effective technique with several advantages; however, its implementation is frequently challenging and requires significant computing power. On the other hand, control strategies based on ε-variables can simplify the control structure, allowing for greater scalability and even resilience. These control strategies are methods that can be utilized to model MG control strategies. The main target of this thesis is to present a hybrid method that can simplify the implementation of MPC using ε-variables and make it more effective for complex energy systems. Our findings indicate that combining MPC with ε-variables can significantly simplify the control structure, thereby allowing the implementation of more complex control strategies. Then, these more complex control strategies can be utilized to provide additional advantages to the energy system, such as scalability and robustness. Although practical methods for modelling control strategies in MGs, ε-variables-based logical control strategies, can make the control structure more scalable, this approach is not optimal. SMPC is an advanced method for controlling power systems while satisfying multiple constraints to achieve an optimal solution based on multiple criteria. Nevertheless, its implementation is complicated. Another target of the thesis proposes a novel systems approach known as the extended optimal ε-variable method, which was created by combining the ε-variable based control method with the S-MPC method in order to address these issues. This novel method has improved the adaptability and scalability of an MG’s control structure and significantly enhanced the MG’s energy management optimization. Our findings indicate that the extended optimal ε-variable method: (i) reduces the operational cost of MG by nearly 35%; (ii) reduces the usage of the battery energy storage system (BESS) by 42%; and (iii) increases the practicability of PV usage by 28%. By translating the results of S-MPC to the ε-variable method, our novel extended optimal ε-variable technique also significantly improves the adaptability and scalability of the control structure of the MG. Regarding flexible hybrid MGs with plug-and-play (PnP) assets, these are difficult to control because of their complexity and asset characteristics. As mentioned previously, ε-variables-based logic control strategies are one method for addressing these challenges. The resulting controller is not, however, optimal. MPC employs a mathematical model of the system to predict its behaviour and determine the best control action. Nevertheless, MPC cannot design and control multiple models, so it cannot control flexible hybrid MGs. S-MPC, on the other hand, uses multiple models to represent system operating modes or scenarios. S-MPC selects a model and control strategy based on the system’s state and performance objectives, enabling it to manage systems with mode-dependent dynamics. However, the development and validation of the design and management of multiple models of the MGs are challenging. Additionally, S-MPC is more challenging to implement due to its multiple steps. Therefore, the last target of this thesis proposes a novel hybrid framework/method based on ε-variables and conventional MPC to generate and validate the S-MPC automatically. To solve the S-MPC optimization problem in a compact form, a quadratic programming (QP) approach that minimizes or maximizes the objective functions under the constraints of bounds, linear equality, and inequality is then considered. This novel strategy significantly improves the energy management optimization of a flexible hybrid MG and reduces computational complexity. The suggested control method is as follows: (i) reduce imported energy from the grid by approximately 46.7% and (ii) increase exported energy to the grid by approximately 50.8%. By translating the decisions of S-MPC into ε-variables, S-MPC implementation is simplified.