DEVELOPMENT OF ENERGY MANAGEMENT STRATEGIES FOR FUEL CELL-BATTERY-SUPERCAPACITOR HYBRID POWER SYSTEM

Abstract

Using renewable energy is becoming a new tendency for vehicular applications to reduce fossil fuel consumption and minimize greenhouse gas emissions. Well-known as an eco-friendly energy source, the proton exchange membrane fuel cell (PEMFC) is extensively used in hybrid power systems to achieve the objective of zero-emission and air protection. However, this type of fuel cell offers slow dynamics and cannot adapt to abrupt load variations when used as a primary energy source. To overcome this shortcoming, battery (BAT) and/or supercapacitor (SC) are supplemented as auxiliary sources. Thus, the hybrid configuration of PEMFC-BAT-SC is developed to improve the performance of the PEMFC system, achieve high operating effectiveness for a hybrid power system (HPS), address the issue of fuel economy, decrease system size, and increase device longevity. In the hybrid power system, it is important to manage the power distribution from energy sources for adapting load power demand under different working conditions. To effectively make use of this hybridization, various energy management strategies (EMSs) have been intensively studied. Essentially, they are classified into rule-based and optimization-based EMSs as two main strategies. Conventionally, rule-based control strategies have been widely used to handle the power distribution of energy sources in real-time. However, these conventional control methods can not be generalized to any optimal objective because the implementation of the rule-based approach significantly depends on the expert knowledge of designers; thus, being not able to fulfill the optimal solution. On the contrary, the optimization-based approach refers to several fundamental algorithms that can specify the optimal objective to enhance the system qualification. Therefore, this thesis presents a comprehensive control strategy to address these inaccurate power distributions, maintain the stability of DC bus voltage, and optimize the economic issues for HPS. Firstly, a comprehensive EMS for an HPS powered by PEMFC, BAT, and SC is proposed with high- and low-level control to improve the accuracy of power distribution from energy sources to the load power demand. In detail, a deterministic-frequency decoupling method is developed in the high-level control to handle the load power adaptation under different working conditions. Then, the low-level control is designed to define correct gains for compensators of current and voltage control loops that maintain the stability of DC bus voltage. The obtained results reveal that the proposed EMS can guarantee power distribution and maintain the robustness of the DC bus under various load scenarios. Next, a high-level control based on fuzzy logic rules and a frequency decoupling method is proposed to improve model uncertainty and complex decisions of deterministic rules. Thexi proposed method ensures a flexible power distribution not only for PEMFC but also for each energy source based on their dynamic characteristics and operating frequency ranges. Next, current and voltage control loops are designed to provide the appropriate gains for compensators that can maintain a stable voltage on the DC bus. As a result, the proposed EMS reduces voltage ripple on the DC bus, while increasing the working efficiency of the PEMFC system. Finally, a novel real-time optimization-based EMS is proposed with a comprehensive structure of high- and low-level controls for a hybrid power system. The proposed methodology is constructed based on an extremum seeking (ES) method combined with a new equivalent SOC and a new adaptive co-state to maintain constraints in the cost function. It is found that the proposed strategy achieves optimal power distribution, minimizes fuel consumption, and improves the PEMFC stack efficiency. All control strategies are validated using the Matlab/Simulink simulations and experiments based on the scaled-down HPS testbench. The obtained results show that the proposed control strategies are both feasible and effective. The final section of this thesis presents conclusions and proposes future studies