Research On Model Predictive Control For V2G Grid-connected Inverter In Electric Vehicle

With the rapid development of electric vehicles,large-scale Electric Vehicle will foces on charging at peak times in the future,which will aggravate the difference between peak and valley load of the grid and increase the burden on the power system.As the energy storage terminal of the future energy internet,intelligentized,gridconnected EV can solve the grid load problem by feedbacking excess power to the grid in the V2 G mode.When the total capacity of the energy storage battery for an EV is large enough,it can provide energy buffering for the grid and reduce the peak-to-valley difference of the grid load.Grid-connected inverter are key equipments for inverting DC power on energy storage batteries into AC power and delivering them to the power grid.Their superior performance can significantly improve grid-connected power quality.Therefore,this thesis develops a study on the control technology for V2 G gridconnected inverters of EV.The main contents of the study are as follows:The overall scheme of V2 G grid-connected inverter was designed.The demand analysis of the inverter was performed to determine the main circuit topology.According to the topology,selecting the appropriate inductance and capacitance parameters,determining the type of the power switch,and design the filter circuit.The V2 G grid-connected inverter model was established.Based on the topology of the inverter,the voltage and current equations of the main circuit are analyzed,the mathematical model in the continuous time domain was established,and the mathematical model of the inverter was discretized to facilitate the implementation of the control algorithm in the digital system.Then,the two parameters of predictive horizon and control horizon were introduced to establish a complete prediction model,which lays the foundation for the design of the predictive controller.The V2 G inverter model predictive controller was designed.Minimizing the current tracking error as the main control target to design the value function.By minimizing the value function,the control variable optimization formula was obtained and the design of the predictive controller was completed.Then,the controller parameters selection was discussed,including the prediction horizon,control horizon,and weight coefficients.The simulation platform was set up to simulate the control effect of the predictive controller.Meanwhile,as a contrast experiment,the traditional PI control was simulated,and the simulation results of the two were analyzed.The prototype of the V2 G grid-connected inverter was designed to realize the practical application of the model predictive control algorithm,and the experiment and result analysis were completed.Combining the control requirements of inverters and the complexity of model predictive control algorithms,a DSP-based control scheme was proposed.According to the design specifications and control requirements of the inverter,the hardware circuit design and software design of the control system were completed.Finally,according to the parameters of the main circuit components and the design of the hardware circuit,building an inverter prototype,completing the model predictive control algorithm and PI control algorithm experiments on the prototype,the experimental results of both were compared and analyzed,verifying the proposed algorithm theory and the feasibility and effectiveness of its implementation.