Topology, Modeling and Control Aspects of DC-DC Converter for an Off-Grid Operated Fuel Cell Power System Perspective

Abstract

With ever increasing demand of electricity, the conventional centralized power system have been steadily replaced by distributed power system (DPS). There are several advantages
of DPS such as modularity, better thermal management, and higher efficiency. DPS finds widespread applications in transportation, distributed generation, consumer electronics and data centers. This dissertation investigates the topological, modeling and controller design aspects of DC-DC converter for an off-grid operated fuel cell power system. In distributed generation system, a renewable energy source is used to harvest utility power. To connect the renewable energy source with a stand-alone load or grid, a power conditioning unit (PCU)is used. A comprehensive review of different DC-DC converter used in PCU has been carried out. Furthermore the classification of DC-DC converters according to the conversion ratio has been carried out. For controller design, an accurate model of DC-DC converter is required. The classical modeling techniques such as state-space averaging modelling and generalized state-space averaging modelling of DC-DC converter provide averaged model
which is suitable only for small-signal analysis. Therefore, parametric system identification technique is used to determine the dynamics of DC-DC converter during the closed-loop operation of the converter. Different linear and block-oriented models and Prediction Error Method are used to estimate the parameters of DC-DC converter in closed-loop. Estimated models are validated using different model validation criteria. Some DC-DC converter exhibits non-minimum phase behavior in their mathematical model. Different methods of mitigating the negative aspect of non-minimum phase behavior has been outlined in this
dissertation. Conventional Type-II and Type-III controller are designed for DC-DC boost converter, but the linear controllers is unable to provide satisfactory performance during large changes in source or load profile due to the fixed-gain nature of the controller. Due to the nonlinear nature of the converter dynamics, nonlinear controller are preferred. One of the most widely used nonlinear controller is sliding mode controller. The classical sliding mode controller has some limitations such as variable switching frequency, steady-state error and chattering effect. Variable switching frequency and chattering effect hampers the practical implementation of the control law. The controller also gives steady-state error which is undesired. To overcome such limitations, fixed-frequency sliding mode current controller and double-integral fixed-frequency sliding mode current controller are designed. In fixed-frequency sliding mode current controller and double-integral fixed frequency sliding mode current controller, the controller provides fixed frequency and removes the unwanted steady state error in output voltage. A multi-staged PCU is developed for fuel cell based stand-alone application which acts as a case study of distributed generation scheme. In fuel cell based stand-alone application, different design constraints such as input
ripple current, decoupling of power stages and power quality have been considered. An active-clamp current-fed isolated full-bridge converter is used as the front-end converter
and single-phase voltage source inverter is used for inverter application. Dual-loop control is used for control of front-end converter whereas fixed-frequency sliding mode controller is used to control the inverter. Simulation results have been provided to validate the design
requirements. From the simulation results it is found out that the said controller gives satisfactory results both in steady-state as well as transient state. The harmonics of the load current also obeys the regulation standard (IEEE standard 519 and IEEE standard 1547). This dissertation provides detailed discussion of topological aspect, modeling and system identification aspect and controller design aspect of DC-DC converter for an off-grid fuel cell power system. A large number of DC-DC converter has been studied and comparative analysis has been provided. Due to the said limitations of state-space averaged modeling scheme, a closed-loop system identification based scheme has been developed. In controller design section, both linear as well as nonlinear controller has been studied and their
performance has been investigated. In the end, an off-grid fuel cell based power system has been designed which can provide power during grid-blackout.