Enhanced control methods for multi-photovoltaic systems and fast/ultra-fast EV charging stations in islanded DC microgrid

With ever increasing of direct current (dc) renewable energy sources and loads, demand for dc microgrid (DCMG) is growing rapidly in urban and particularly in remote areas where there is a better access to dc distributed energy resources (DERs). DCMG can be connected or isolated from the main grid to operate in either grid-connected or islanded mode, respectively. Islanded DCMG specifically offers several advantages for both users and electricity providers in remote areas where power transmission is costly or unfeasible. For example, it provides higher efficiency, more reliability and better controllability.

Despite of the benefits of islanded DCMG, providing smooth and reliable electricity for the users is always a challenge. In islanded DCMG, regulation of the dc-bus voltage is not as fast as regulation in grid-connected DCMG. This could be due to the low inertia nature of the interconnecting converters, which slows down the regulation process, while creating more oscillations on the dc-bus voltage. This issue is more critical during high variations on renewable energy sources' output power or at fast switching of bulk loads such as fast/ultra-fast electric vehicle (EV) chargers.

The intermittent nature of solar irradiation and wind speed makes the output power of renewable energy sources (i.e. photovoltaic systems (PVs) and wind turbines) be fluctuating. As a result, the dc-bus voltage experiences oscillations that can be harmful for some sensitive loads. Although, battery energy storage system (BESSs) is a common solution addressed in the literature, it cannot satisfactory reduce the oscillations of the dc-bus voltage due to the low inertia of the BESS's converter as well as time-to-time absence of the BESS when it is fully charged or flat.

Moreover, even during the presence of the BESS, DCMG is still vulnerable against connection/disconnection of bulk loads such as fast/ultrafast EV charging stations. Each fast/ultrafast EV charger demands around twenty times of a typical house consumption. This creates fast and high transient on the dc-bus voltage during start and stop time of EV charging.

To overcome these challenges and limitations, novel and advanced control strategies, as well as well-developed power management systems are proposed in this thesis. The proposed control strategies are aimed for the interlinking converters of the DERs and the fast/ultra-fast EV chargers in islanded DCMGs. The three main contributions of this thesis are detailed as follows:  

The first contribution is to design a distributed dc-bus signalling control approach for PVs in an islanded DCMG. The proposed control method makes the islanded DCMG less dependent on the existence of the BESS. In this method, the dc-bus voltage is monitored and compared with a predefined value. Based on the obtained deviation and the voltage derivative, the operating mode of all the PVs automatically switched between maximum power tracking mode and voltage regulating mode. Such switching results in protecting the DCMG from unavoidable shutdown that is conventionally necessary during the absence of the BESS unit. Moreover, the proposed control method reduces the oscillations on the DC-bus voltage during the availability of the BESS.

The second contribution is to design an optimized distributed cooperative control method for multi-photovoltaic energy sources in an islanded DCMG. This control method minimizes the impact of the intermittent nature of PVs on the total generation and reduces the charging/discharging fluctuations of the BESS. By using the proposed controller, a constant and controllable total generated power is obtained from all PVs that is costly optimized. In addition, more energy is stored in the BESS compared to the conventional methods that can support the DCMG during the absence of sunlight or when the PVs are in standby mode.  

The final contribution is to propose two enhanced control methods for an ultra-fast EV charging station in an islanded DCMG. The proposed approaches are software-based solutions and significantly reduce the undershoot/overshoot of the dc-bus voltage caused by the ultra-fast charging process. Unlike conventional hardware-based charging methods, which require large and costly such as using supercapacitors and flywheels to damp the oscillations of the bus voltage, the proposed methods do not require these types of equipment. Moreover, the proposed controllers are robust against unpredictable disturbances.

The dynamic models of all proposed control methods are thoroughly analyzed, and their effectiveness and superior performance compared with the conventional methods are validated through different case studies in MATLAB/Simulink.