Research On Robust Control Of DC Microgrid System With Photovoltaic And Energy Storage Based On Model Predictive Control

In recent years,the demand of people for electricity is growing,and the problem of ecological environment is becoming more and more serious.With the goal of " Dual Carbon" was proposed,the greatly development of clean energy has become an effective way to build an energy-saving and environment-friendly society.The emergence of microgrid technology provides technical support for the local consumption,rational allocation and safety management of green renewable energy.Owing to its simple topology and without considering reactive power and phase synchronization,in recent years,the DC microgrid has become a hotspot for research.Firstly,the thesis takes the photovoltaic(PV)and energy storage system(ESS)as the major research object.in which the PV and ESS are used as the main power supply units.The PV and ESS are modelled mathematically and the corresponding controllers are designed for the PV and ESS respectively.The model predictive control(MPC)is based on the minimization of a predefined cost function by studying the predicted response of a power converter over a finite time duration at each time step.Due to its fast dynamics and flexible control scheme in which different constraints can be readily formulated,MPC has been widely used for the control of power converters.Therefore,in order to improve the robustness of the PV and ESS DC microgrid system,model predictive control algorithms,without any PID controller,are designed for the PV maximum power point tracking(MPPT)and constant voltage control(CV)modes and for the DC converter of the energy storage system respectively.The specific research contents of this thesis are as follows:(1)According to the topology of the typical PV and ESS DC microgrid system studied in this thesis,firstly,the corresponding mathematical models are established for PV cells,PV system,battery,supercapacitor and energy storage system respectively.Secondly,two operating modes of PV in different situations are analysed,i.e.MPPT mode and CV mode.Finally,the power balance of the DC microgrid system is analysed.(2)Aiming at the problems of the traditional P&O method can’t guarantee the dynamic performance and steady-state accuracy at the same time when the external environment changes,for PV system,an improved P&O method MPPT control strategy based on MPC is proposed.Firstly,the motion of PV operating point is divided into forward motion and reverse motion,and based on this,the perturbing step is divided into large and small step sizes.The forward motion corresponds to small step and the reverse motion corresponds to large step,so as to reduce the oscillation of the PV at the maximum power point(MPP)and increase the dynamic performance and robustness of the PV system for external environmental changes.Secondly,by analyzing,it can be known that the impact of solar irradiance on PV output power is much greater than that on PV output voltage and current.Therefore,the changing value of PV output power is used to determine whether the solar irradiance changes.If the solar irradiance changes,the MPPT control reference current will be modified to furtherly improve the dynamic performance of the PV system.And a small step perturbing step is taken in order to reduce the impact of operating point drift on the system,Based on this the robustness of the PV system is further enhanced.Finally,the method designed in the thesis is compared with the conventional P&O control with fixed,variable and adaptive step sizes to verify the feasibility and effectiveness of it.(3)Aiming at the problems of the energy storage system is taken out of operation due to overcharging or faults in the PV and ESS DC microgrid system,so that it can not to ensure constant DC bus voltage,an MPC PV constant voltage control(MPVC)strategy based on an extended state observer(ESO)is proposed.Firstly,the optimal operating region is determined,by analysing the two different operating regions of CV mode of the PV system.Secondly,the relevant ESO is designed to estimate the PV system partial quantities online,reducing the use of relevant sensors and the system cost.Thirdly,the MPVC controller was designed to achieve constant control of the DC bus voltage.Finally,the fast dynamic performance and strong robustness of the method porposed in the thesis,by comparing with the conventional PI control strategy,are verified from both simulation and experiment by using MATLAB /Simulink and the FPGA-based Star Sim semi-physical simulation platform.(4)For the problem of the PI parameters rectification complex about hybrid energy storage system PI control method,an MPC-based hybrid energy storage control strategy is designed.Firstly,through the second-order low-pass filter divide power,which need energy storage system to balance,into low-frequency and high-frequency power.Secondly,the MPC-based battery control strategy and supercapacitor control strategy are designed to compensate or absorb the low-frequency and high-frequency power respectively,in order to achieve the aim of balancing power and stabilize the DC bus voltage.Finally,the system model is built in the simulation platform to simulate the operating conditions under different situations in practice to verify the effectiveness of the designed control strategy.