Modelling and control of modular multilevel converter for high-speed railway traction power supply

Over 50 countries and regions worldwide have communicated their commitments to net-zero by the middle of this century. Transportation is responsible for 27% carbon emissions in the UK in 2019.Railway is recognised as one of the most energy efficient means for travel as well as for freight transport. However, only 38% of the current railway route is electrified in the UK. Diesel-based internal combustion engine trains will be phased out in the next decades. Long term plans are necessary for the success in meeting the 2050 goal. This includes not only electrifying the non-electrified railway network, but upgrading the existing infrastructure to increase power supply capacity. The traditional feeder stations in traction power supply systems (TPSS) are based on different transformers to convert power from the three phase utility grid power to a single phase traction AC overhead line. Power quality issues such as voltage imbalance and current harmonics exist in the legacy system, and the neutral sections between each supply phase hinder the evolution of the traction power network to a fully connected smart network. Furthermore, this type of station does not provide convenient interfaces for renewable source integration. The electronic converter based static frequency converter (SFC) has been introduced into the feeder station construction. This SFC solution can solve most of the power quality issues through its flexible operation scheme, and RES can be integrated into the SFC based feeder station by altering the converter topology. The modular multilevel converter (MMC) is viewed as a promising topology for high power conversion applications. However, the cost of this solution is high at the moment when using SFC to replace the station transformer. In this thesis, a MMC is chosen as the primary topology and novel control strategies are developed to improve its performance. Firstly, the traditional cascaded linear control approach is summarised and modified to improve the DC power delivery performance. Then model based MMC modelling and its detailed internal dynamics are analysed for long horizon predictive controller design, which provides superior transient response speed over conventional approaches. Regarding railway TPSS, a compensation method for voltage imbalance issue in the V/v transformer substation is investigated. A simulation platform for AT-based TPSS is then developed to support mobile traction load analysis under different supply control schemes. To mitigate the power quality issue with reduced investment, a partial compensation strategy based cophase scheme is investigated. The proposed design and control method reduces electronic converter capacity, which provides a hybrid approach for substation design with lower SFC investment. To improve the SFC operational performance, a long horizon model predictive controller is designed for back to back MMC based feeder station, which has optimal performance under the most demanding load condition. To integrate the renewable energy generation, wind generation power is coupled at the DC-link of the back to back MMC substation. A modified MMC control is designed for DC link voltage stability despite drastic generation/load change. To improve substation reliability and energy efficiency, batteries are interfaced to each submodule in the MMC arms. This topology provides greater connection flexibility and operational reliability. The internal energy for charging and discharging is controlled by MPC under the hierarchical scheme. The ESS can be charged by utility grid or a train's regenerative power, and can be discharged for supporting traction load in grid fault condition. To fully utilise the freedom of SFC substation control, a power sharing strategy is proposed in the two SFC stations scenario where two SFCs are collaboratively supplying an AT-network from both ends. A frequency droop method is designed to assign the optimal power to each SFC. The proposed method achieves balance in capacity enhancement and transmission loss reduction. In summary, this thesis presents a comprehensive study on the modelling and control of MMC based SFC applications for future railway power suppl. The proposed work has the potential of reduction SFC overall cost by smaller capacity converter and smaller size capacitors. Renewable sources integration approaches may assist with transforming current TPSS to an eco-friendly and smarter system in the future.