Research On Voltage Equalization Strategy And Implementation Of Charging System For Supercapacitors On Energy Storage Type Urban Rail Transport

As an emerging energy storage material, supercapacitor1has many advantages over energy storage batteries, such as high power density, quick charge/discharge and more cycle times. It has been a research focus to use supercapacitors as power supply for urban rail vehicles. In industrial practice, supercapacitors are often connected in series and parallel to form a supercapacitor stack, due to a supercapacitor’s low endurance voltage, despite its large capacitance. Meanwhile, capacitors on the vehicle must be charged during the short stopping time. In this dissertation, the voltage equalization2and fast charging issues of supercapacitors is considered, and then a novel charging system is developed.The supercapacitor stack on urban rail vehicles are connected in series and parallel by thousands of supercapacitors, so it is necessary to consider the voltage equalization problem of the capacitors. The efficiency of the super-capacitor stack is influenced greatly by the voltage equalization system. A novel series-parallel connection method is proposed to achieve the voltage equalization of thousands of supercapacitors. Considering the voltage equalization of switched capacitors with a hierarchical voltage-sharing topology, a switched capacitors control based on cooperative control algorithm for supercapacitors’voltage equalization is proposed, and stability of the algorithm is proved by Lyapunov function.The main circuit of supercapacitor stack charging system is designed and parameters are verified to satisfy the charging demand of energy storage electric urban rail vehicles. The input AC10kV is rectified to produce the DC-link for the main circuit and the DC/DC converter is realized by PWM chopped wave. Then constant current charging can be realized by controlling the PW pulse width. Two branches of Buck circuits are in parallel to achieve the reliability. Then the harmonics of charging current will decrease by control of parallel interleaved Buck DC/DC Converter.Model Predictive Control (MPC) strategy is adopted to optimize the charging system. The charging model of supercapacitors and physical model of the main circuit are constructed first. Then the model is modified by normalization method, and the system model is discretized. Considering the current reference of the charging system as the optimization performance, a finite horizon optimization problem can be constructed with some hard constraints, such as duty ratio, current peak value of inductors. Then the model predictive control algorithm is proposed and the charging system is optimized and adjusted.The proposed scheme is verified in a united simulation platform constituted with Saber and Matlab/Simulink. The circuit model is simulated in Saber, while the model predictive control algorithm is realized in Matlab/Simulink. With the co-simulation platform, the voltage sharing strategy, the main circuit of charging control strategy with model predictive control algorithm are verified, current sharing is relized and the time of charging is reduced in this dissertation.The structure of charging system is proposed and the hardware, software and debugging work has been completed. The MPC based charging system has been applied to the first civil urban rail vehicle, which demonstrate the effectiveness of the proposed scheme.